Argument

Evidential Problem of Evil Defeater

Intro

If God is real, why does the world contain so much pointless suffering? Not just bad things in general, but suffering that looks like it serves nothing, a fawn burning slowly to death in a forest no human will ever walk through, a child dying of cancer with no growth or witness to redeem it. That is the evidential problem of evil, the strongest version of the argument from suffering. It does not claim God and suffering are contradictory. It claims the kind and amount of suffering we actually see makes God's existence unlikely.

The objection is honest, serious, and personal. It is not a debater's trick. Many people, including some former Christians, leave faith because they cannot square the world they live in with a loving God. Any reply that does not feel the weight of that has not understood the argument.

The Christian response runs on four layers, and they reinforce each other. First, the inference at the heart of the argument, "I cannot see a good reason for this suffering, therefore probably no good reason exists," is the kind of move that fails reliably elsewhere. A beginner watching a chess grandmaster cannot see the point of a sacrifice; that does not mean the sacrifice is pointless. Finite human minds against an infinite mind face a similar gap. Second, the Christian framework actually supplies candidate reasons. Free moral creatures require real consequences. Character forms through struggle. The cross is God's own entry into human suffering, not a distant audit of it. The end of the story includes a wiping away of every tear. Third, the argument itself depends on objective moral standards (that gratuitous evil is really evil); naturalism cannot ground those standards, so the argument quietly borrows from theism while attacking it. Fourth, the hardest cases (natural disasters, animal suffering before humans existed) get their own targeted responses.

The Christian is not claiming to know why any specific evil happens. The claim is more modest: the argument's logic, from "I see no reason" to "no reason exists," does not survive scrutiny against a being whose vantage point is, by hypothesis, vastly beyond ours. And it does not match the world the New Testament actually describes, a world where God Himself enters the worst of it and bears it.

This page lays out the full four-layer defense in debate-prep form, with the objections, rebuttals, and tactical notes apologists like Craig, Plantinga, and Stump have field-tested for decades.

In full

The evidential problem of evil, the claim that the amount, kinds, and apparent gratuitousness of suffering in the actual world makes God's existence improbable, is the live battlefield of modern apologetic engagement after Mackie's logical POE collapsed (per Problem of Evil, Free Will Defense and All-Do-Good World POE Defeater). The canonical atheist deployment is William Rowe's "fawn in the forest" (The Problem of Evil and Some Varieties of Atheism, 1979): a fawn dies slowly of burns from a lightning-strike forest fire, unseen by any human, with no apparent moral payoff. Probably, Rowe argues, no God-justifying reason exists for permitting such evil, so probably no God exists. This page is structured as debate prep for the deployments William Lane Craig used against Walter Sinnott-Armstrong (God? A Debate, 2004) and Bart Ehrman (March 2006), Lennox used against Hitchens / Dawkins, and Plantinga formalized in Where the Conflict Really Lies (2011). The single most important live-debate move is distinguishing the Skeptical Theism inferential block from the positive-theodicy composite, these are complementary moves operating at different levels (epistemic and metaphysical), and confusing them is the most common apologetic-failure mode. Companion to Problem of Evil, Free Will Defense (master defense, the logical POE), All-Do-Good World POE Defeater (Angstreich's modal-polished logical POE), and Skeptical Theism (the load-bearing concept hub for D1 below). Note: Jack Angstreich does not surface as a deployer of the evidential POE in his ingested corpus, his arena is the logical POE, but Christian apologists who beat Angstreich on the logical move should expect the next atheist to shift to evidential ground, and this defeater is the standing prep for that shift.

Argument structure

The atheist's argument (Rowe-style evidential POE):

The dialectical force is inductive / probabilistic: Rowe does not claim God's existence is contradicted by evil (Plantinga answered that, see Problem of Evil, Free Will Defense); he claims the actual kinds and amounts of evil make God's existence improbable. The argument hinges on the inferential move from we see no God-justifying reason for E to probably no such reason exists.

The Christian defense (4-layer composite):

Form

Defensive composite: inferential-block + positive theodicy + meta-ethical reductio + cosmological sub-defense. Unlike the Problem of Evil, Free Will Defense (which is purely a possibility-claim against the logical POE), the evidential POE defense is substantive, it makes positive claims about God's reasons, the structure of moral reality, and the cosmological framework. The four layers are complementary, not exclusive; the apologist deploys whichever combination is most live in the specific debate. Modern field-tested debates show: skeptical theism is the universal anchor (D1); Craig pairs it heavily with the moral-grounding move (D3); Lennox pairs it with the Christological theodicy (D2); academic engagements (Stump, M. Adams, Bergmann) emphasize the eschatological / Christological dimensions of D2.

D1, Skeptical theism: the inferential block

Affirmative case (second-order arguments)

1. The CORNEA principle (Wykstra 1984) blocks Rowe's inductive move. Condition of Reasonable Epistemic Access: an inference from "p appears to be the case" to "p is probably the case" is reasonable only if it is reasonable to believe that, if not-p were the case, this would have shown up given the observer's access. Applied to POE: the move from "I see no God-justifying reason" to "probably no such reason exists" is reasonable only if "if such a reason existed, it would likely have shown up to my finite cognitive faculty." But the human-divine cognitive asymmetry makes this assumption implausible, most God-justifying reasons (if any) would involve considerations across vast temporal-causal chains, counterfactual ramifications, and metaphysical relations that finite cognition cannot reliably access.

2. The chess-novice analogy (Wykstra). A chess novice watching a grandmaster's move that looks pointless: the novice's failure to see the move's purpose is no evidence that the move has no purpose. The novice lacks the chess-cognitive resources to detect the purpose if one is present. Replace "chess novice" with "human finite intellect" and "grandmaster" with "God omniscient over all causal-counterfactual chains": the same logic blocks the Rowe inference. This is the most-used analogy in field-tested deployments.

3. The track-record argument (Alston 1991). Human moral judgment about distant or large-scale matters has a poor track record: 19th-c. moral judgment about race; medieval moral judgment about heresy-execution; Greco-Roman moral judgment about chattel slavery. In each case, the considered moral judgment of the period was strongly held and demonstrably wrong. If finite moral judgment is this unreliable about human-scale moral matters, it is even less reliable about cosmic-scale moral matters where God's vantage encompasses considerations no human has access to.

4. The probability-theoretic argument (Bergmann 2006). Even granting Bayesian frameworks, the prior probability that we would identify all morally-sufficient reasons God might have is low given the cognitive asymmetry. The evidential POE's posterior probability against theism is correspondingly weak, failure to identify reasons does not raise atheism's probability significantly. This is the technical-philosophical version of the Wykstra move.

Anticipated objections

1. "Skeptical theism is just God-of-the-gaps for ethics." The atheist re-frames: the theist appeals to unknowable divine reasons to bridge an explanatory gap, the same illegitimate move as appealing to unknowable divine causes in nature.

2. "Skeptical theism collapses into wider moral skepticism." (Maitzen 2009; Almeida & Oppy 2003.) If we can't trust our moral judgments about cosmic-scale suffering, can we trust any moral judgment? The position seems to undermine ordinary moral reasoning.

3. "The Job-paradigm is intellectually lazy." Skeptical theism's biblical anchor (Job 38-41, God doesn't justify the suffering but displays the asymmetry) is dismissed as a non-answer that wraps inscrutability in religious-rhetorical packaging.

Rebuttals

1. The God-of-the-gaps charge misfires because skeptical theism is not an explanatory move. A "God-of-the-gaps" hypothesis posits unknown divine action as a positive explanation for a gap in natural-causal understanding (e.g., "we don't know how lightning works, so God did it"). Skeptical theism is the opposite: it does NOT posit any positive divine reason for any particular evil, it only blocks the atheist's inferential move from "no apparent reason" to "no actual reason." The theist who deploys skeptical theism is making no positive existence-claim about hidden reasons; they are challenging the warrant of the atheist's negative existence-claim. The two argumentative positions are not parallel. Failure mode: confusing a defensive epistemic move with an explanatory metaphysical move.

2. The moral-skepticism collapse charge depends on a non-existent slippery slope. Skeptical theism is a local epistemic move about a specific class of judgments, namely, judgments about whether God could have justifying reasons we cannot access. It is not a global skepticism about all moral judgment. The position holds: we can confidently judge that gratuitous cruelty is bad (in the local human sense) while failing to confidently judge whether every instance of apparently gratuitous cruelty is in fact gratuitous given divine vantage. The Bergmann response makes the local-vs-global distinction technical (the principles of skeptical theism are restricted to a class of inferences, not universalized). Failure mode: confusing local epistemic modesty with global moral skepticism.

3. The Job-paradigm is not inscrutability-wrapped-in-rhetoric, it is the canonical biblical articulation of the cognitive-asymmetry response. Job 38:4: "Where were you when I laid the foundations of the earth?" Job 40:8: "Will you condemn me to justify yourself?" The text does not say the suffering had no reason; it says Job is not in a cognitive position to judge that it had none. This is exactly the skeptical-theist structure, articulated 2,500 years before Wykstra. Far from being a non-answer, the Job-paradigm is the paradigmatic answer, it is what skeptical theism is doing in philosophical-technical dress. (See Skeptical Theism §"The Job-paradigm" for the developed treatment.) Failure mode: dismissing a sophisticated epistemic move as religious-rhetorical evasion.

Live-cite kit

• Scripture: Job 38-41 (God's response to Job's complaint); Romans 11:33-36 ("Oh, the depth of the riches both of the wisdom and knowledge of God! How unsearchable are His judgments and unfathomable His ways!"); 1 Corinthians 13:12 ("now we see in a mirror dimly").
• Scholarly: Stephen Wykstra, "The Humean Obstacle to Evidential Arguments from Suffering" (International Journal for Philosophy of Religion 1984, the foundational CORNEA paper); William Alston, "The Inductive Argument from Evil and the Human Cognitive Condition" (Philosophical Perspectives 1991); Michael Bergmann, Justification without Awareness (2006, ch. 5); Daniel Howard-Snyder & Michael Bergmann (eds.), Skeptical Theism: New Essays (2014); see Skeptical Theism for the full reference apparatus.
• Aphorism: "The chess novice doesn't refute the grandmaster by failing to see the point of the move."

Tactical notes

• Lead with Wykstra's chess-novice analogy. It is the single most accessible illustration of the cognitive-asymmetry point. Audiences, even philosophically unsophisticated ones, grasp it immediately. Craig deploys variants of it almost every time the evidential POE comes up in debate.
• Don't defend skeptical theism as the only response. It is the first line, the epistemic block. Pair it with D2 positive theodicy so the position doesn't look purely defensive. The "fortress + sortie" model: D1 blocks the attack; D2 demonstrates positive content.
• Force-commit move: "Are you claiming that you can positively identify that no God-justifying reason exists for [horrible case E]? If yes, on what cognitive grounds? If no, the inferential move from 'I see no reason' to 'no reason exists' is what we're disputing."

D2, The positive-theodicy composite

Skeptical theism (D1) is defensive, it blocks the atheist's inference. Modern field-tested debates pair it with a positive theodicy composite that names candidate reasons God might have for permitting evil. The composite has three primary layers, each over-determined.

D2a, Christological theodicy

The strongest theological-and-debate-tested response. The cross is not a defense of God's permitting evil from a distance; it is God's voluntary entry-into suffering. The Incarnation makes God a participant in human suffering rather than an exempt overseer. The eschatological climax (resurrection) defeats death from within rather than by external decree.

Key moves:

1. Christ's entry-into-suffering changes the moral framing. Marilyn Adams (Horrendous Evils and the Goodness of God, 1999): horrendous evils, life-ruining-in-themselves evils, cannot be compensated by greater goods in a consequentialist calculation. The proper theistic response is Christological: God enters horrendous suffering in Christ, and the eschatological consummation makes the sufferer's life integrated with Christ's suffering, not just compensated for. This is the most-deployed academic response to Rowe-style horrendous-evil cases.

2. The cross is the dispositive theistic act in the face of evil. Eleonore Stump (Wandering in Darkness, 2010): the response to the POE is not propositional defense but narrative, the Christian narrative of God's response to suffering centers on the cross and resurrection. The narrative reveals what propositional argument cannot capture: that God's love is not absent in suffering but enters it.

3. Live-cite force: when Bart Ehrman in his 2006 debate vs Craig pressed the evidential POE personally (his deconversion narrative was POE-driven), Craig's response repeatedly invoked the cross as God's solidarity with suffering, not as theodicy-substitute but as theological context that re-frames the question.

Limitations of the move:

• The Christological theodicy works within a Christian framework. Against a fully naturalist atheist who rejects the Incarnation as historical claim, it functions as a consistency-claim (the Christian worldview internally answers the POE) rather than a converting argument.
• It does not address natural evil prior to the Incarnation (3.5 billion years of animal suffering before Calvary), that load lands on the natural-evil sub-defense (D4).

D2b, Eschatological theodicy (compensation + telos)

The temporal/finite scale of suffering is dwarfed by the eternal/infinite scale of redemption.

Key moves:

1. 2 Corinthians 4.17: "For our light momentary affliction is producing for us an eternal weight of glory beyond all comparison." The Pauline framing makes suffering instrumentally productive of eschatological glory, not merely compensated by it, but generative of it.

2. Romans 8.19-22: "the creation itself will be liberated from its bondage to corruption and obtain the freedom of the glory of the children of God." The cosmic narrative arc moves from groaning creation through Christ's redemption to consummated glory. Suffering is a phase in a redemption-story, not a permanent feature of reality.

3. Revelation 21.4: "He will wipe away every tear from their eyes; and there will no longer be any death; there will no longer be any mourning, or crying, or pain." The eschatological telos defeats evil, not by retroactively justifying it, but by terminating it.

4. Craig's debate-tested deployment: in the Sinnott-Armstrong debate (2004), Craig consistently pressed that an eternal-perspective changes the moral calculus on actual evil. The fawn-in-the-forest case looks horrific from a sub-specie temporis vantage; from an eternal vantage that includes eschatological compensation for all sufferers (including the fawn, Romans 8:19-22 includes the whole creation), the calculus changes.

Limitations:

• The eschatological response works for victims who are redeemed; it does not address the eternally damned (the population for whom suffering is not eschatologically compensated). That objection sits in Hell and Eternal Punishment, a separate apologetic-load, and is partly answered by the move that the damned have chosen their state via libertarian free will.
• Critics (Marilyn Adams) press that some horrendous evils are so life-ruining that no compensation-in-the-future can integrate them. Adams's response is to push to the Christological move (D2a) rather than rest on pure compensation, the integration of the sufferer's life into Christ's life is what works for horrendous evils, not mere addition of future goods.

D2c, Soul-making theodicy

John Hick's Evil and the God of Love (1966) developed the Irenaean alternative to the Augustinian theodicy. Suffering serves the formation of moral and spiritual character. Courage, patience, compassion, perseverance, sacrificial love, these virtues cannot exist without the existence of trial. A world without suffering would be a world without these specific goods.

Key moves:

1. Specific virtues are response-dependent on specific evils. Compassion presupposes someone to be compassionate toward; courage presupposes danger; patience presupposes adversity. A world without evil is a world without these particular forms of goodness. The atheist must show that this trade-off is negative on net, not just that evil is bad in isolation.

2. The hedonic-paradise critique. Hick: a world of constant pleasure without challenge would be a hedonic paradise, but not a moral paradise. The world that produces characters is one with friction. This re-frames the question: not "could God create a less painful world?" but "could God create the same kind of character-formed creatures in a less painful world?"

3. Stump's narrative deepening: Hick's framework is sometimes criticized as making suffering instrumentally-valuable in ways that feel callous. Stump's narrative-Christological theodicy preserves the soul-making structure while embedding it in personal-relationship terms, suffering forms relationship with God and integration with Christ, not mere virtue-tokens.

Limitations:

• The soul-making theodicy struggles with apparently gratuitous suffering (suffering that doesn't lead to growth, extinction-level events, animal suffering with no moral subject to form). Hick's own response: the eschatological hereafter completes the formation that earthly suffering started. This re-loads onto the eschatological theodicy (D2b).
• It struggles with animal suffering (animals don't form moral characters in the human sense). That load lands on the natural-evil sub-defense (D4).

The composite move

In live debate the three theodicies are deployed together, not exclusively. The atheist who attacks one theodicy must reckon with the others; defeating one does not defeat the composite. Craig's standing deployment uses Christological + eschatological as primary, with soul-making as supplementary; Lennox uses Christological-eschatological-soul-making roughly equally; academic engagements (Stump, M. Adams) emphasize the Christological-narrative dimension. The composite is over-determined, the atheist's job becomes much harder.

Anticipated objections to the composite

1. "The Christological move is just smuggled exclusivism." The atheist objects that invoking the Cross presupposes Christian theism, which is the position under challenge.

2. "Eschatological compensation is a cosmic IOU that can't be verified." The atheist presses that promising eternal goods to compensate temporal suffering is empirically unverifiable.

3. "Soul-making fails for gratuitous cases." Some suffering produces no growth; some sufferers are too damaged to grow; animal suffering can't be character-formative.

Rebuttals

1. The composite is dialectically fair, atheist arguments routinely smuggle naturalism (an entire worldview) into evidential-POE deployments. When the atheist says "the actual amount of evil makes God's existence improbable," they presuppose that naturalism (no God) is the comparison-case and that what we observe counts as evidence. The Christological move is the symmetrical theistic deployment: within the theistic framework, here is the response. The atheist cannot demand the theist argue without invoking theism while the atheist freely invokes naturalism. Failure mode: asymmetric methodological demand.

2. The eschatological-compensation move is no more empirically-unverifiable than the atheist's grounding claims. Both sides are making metaphysical claims about ultimate reality that go beyond the empirically-verifiable. The atheist who deploys evidential POE is making claims about the moral weight of evil (an unobservable normative fact) and about what a hypothetical God's reasons would be (counterfactual reasoning about a being whose existence is at issue). The theist's eschatological move is dialectically symmetrical. The "unverifiable" charge defeats both positions or neither. Failure mode: selective epistemic standards.

3. The "gratuitous cases" objection re-runs the Rowe argument at a sub-layer, which is exactly what skeptical theism (D1) is for. When the atheist points to "this specific case has no compensating growth, no character-formation, no integration with anything good", they are making the same inferential move from "no apparent compensating good" to "no compensating good exists." D1 blocks this move at the local-case level just as at the global level. The composite holds: D2 names positive candidate reasons that the atheist hasn't refuted; D1 blocks the atheist's negative existence-claim about candidate reasons.

Live-cite kit

• Scripture: 2 Corinthians 4.17 (light momentary affliction → eternal weight of glory); Romans 8.19-22 (creation groans → liberation); Revelation 21.4 (every tear wiped); 1 Peter 1:6-7 (trials prove faith more precious than gold); James 1:2-4 (testing produces endurance, character, hope); Hebrews 5:8-9 (Christ learned obedience through what he suffered).
• Scholarly: Marilyn McCord Adams, Horrendous Evils and the Goodness of God (1999); Eleonore Stump, Wandering in Darkness: Narrative and the Problem of Suffering (2010); John Hick, Evil and the God of Love (1966); Alvin Plantinga, Where the Conflict Really Lies (2011, ch. 4-5); William Lane Craig, Reasonable Faith (3rd ed. 2008, ch. 4); Richard Swinburne, Providence and the Problem of Evil (1998).
• Aphorism: "The Cross is the dispositive theistic act in the face of evil, not a defense, but a presence."

Tactical notes

• Christological theodicy is the strongest live-debate move. In Craig vs Ehrman 2006, this is what Craig kept pressing, and what visibly disarmed Ehrman, whose deconversion narrative was POE-driven and who had no atheist analogue to "God enters suffering himself." When the atheist's argument has a personal-suffering anchor, Christological theodicy is dialectically and pastorally devastating.
• Eschatological compensation is rhetorically powerful but academically contested. Use it confidently in popular debate; expect Marilyn-Adams-style horrendous-evils pushback in academic engagement.
• Don't lead with soul-making. It's the weakest of the three for live debate (the "gratuitous cases" objection lands hard against it). Soul-making works supplementarily, once Christological and eschatological have done the work.

D3, The grounding pin: evidential POE presupposes objective moral standards

Rowe's argument requires the moral premise (A2): an omnibenevolent being would prevent certain kinds of evil. This premise depends on the existence of objective moral standards by which evil is judged. The atheist must therefore answer: where do those moral standards come from?

Affirmative case (second-order arguments)

1. The argument is internally incoherent on naturalism. If naturalism is true, there are no objective moral facts. There are only evolved human preferences and social conventions. "X is gratuitously evil" reduces to "humans typically dislike X", but that is not the kind of moral fact the evidential POE requires. The argument requires objective normative standards by which we measure God's permitting of evil. Naturalism cannot supply those standards. The argument is therefore borrowing moral realism from theism while attacking theism. This is Craig's signature deployment in his Sinnott-Armstrong debate (2004) and many subsequent engagements.

2. The atheist's response must specify a non-theistic grounding for moral realism, and the options are weak. Wielenberg's Robust Ethics (2014) attempts a non-theistic moral realism via brute moral facts (Platonic-style). Critics (Craig, Copan) press that brute-fact moral realism is metaphysically extravagant on naturalist parsimony grounds, and that the connection between physical-natural facts and moral facts is unexplained. Other naturalist options (constructivism, expressivism) make morality non-objective, which makes the evidential POE argument toothless, "I really really dislike X" is not "X is objectively evil."

3. C.S. Lewis's "puddles of water" line: a puddle of water, if it could speak, might think the hole it occupies is perfectly fit for it. But the hole was not made for the puddle; the puddle was shaped to the hole. Our moral perception that gratuitous suffering is bad tracks something objective, the trackings of finite moral perception to moral reality is what theism explains and naturalism cannot. To use the moral perception as evidence against the theism that explains it is to saw off the branch one sits on.

Anticipated objections

1. "Atheists can have moral standards without God, the Euthyphro Dilemma shows theistic grounding fails." The standard objection: theistic grounding of morality faces its own problem (the Euthyphro), and atheists can ground morality via reason, social contract, or evolutionary preference.

2. "You're conflating moral epistemology with moral ontology." The atheist may agree that knowing what is moral does not require God, even if being moral requires God. The argument doesn't depend on objective moral facts in the strong sense.

3. "Moral realism is independent of theism, secular philosophers (Parfit, Wielenberg) defend it." The non-theistic moral-realism literature is real and growing.

Rebuttals

1. The Euthyphro Dilemma has been answered by classical-theistic ethics (Aquinas, Adams). God's nature is the grounding of moral goodness, God is not subject to an external standard (Euthyphro Horn 1) nor arbitrarily declaring standards (Euthyphro Horn 2). God's nature is the moral standard. The dilemma collapses on the classical-theistic ontology of God's nature as identical with the moral good (per Divine Simplicity). (The Euthyphro Dilemma deserves its own hub, held as plain text per ghost-discipline; Plato's Euthyphro 9d-11b is the locus classicus; classical-theistic responses in Augustine, De Libero Arbitrio II.10; Aquinas, ST I-II q. 91 a. 1; Robert Adams, Finite and Infinite Goods 1999.) Failure mode: deploying the Euthyphro Dilemma without engaging classical-theistic responses.

2. The epistemology/ontology distinction doesn't save the argument. If the atheist concedes there are objective moral facts (so the evidential POE has teeth) but claims they exist independently of God, they have endorsed moral Platonism, a metaphysically extravagant view that naturalism's commitment to parsimony cannot easily accommodate. The naturalist who is also a moral realist is in tension with their own naturalism. The argument forces the atheist to choose: surrender moral realism (and toothless POE) or surrender naturalism (and conceded theistic-style moral metaphysics). Failure mode: thinking objective morality is metaphysically free.

3. Wielenberg-style non-theistic moral realism is contested, not settled, and faces the connection-problem. Robust Ethics posits brute moral facts that supervene on natural facts, but the supervenience relation is unexplained. Why does this natural configuration ground this moral fact? On theism, the answer is divine teleology and divine intellect; on Wielenberg's view, the supervenience is brute and unexplained. The theistic ontology has the better explanatory profile. Failure mode: assuming the contested philosophical position is established.

Live-cite kit

• Scripture: Romans 2.14-15 (the law written on hearts); Genesis 9:6 (the imago-Dei grounding of murder's wrongness); Psalm 19:1 (heavens declare the glory, natural revelation includes moral structure).
• Scholarly: William Lane Craig vs Walter Sinnott-Armstrong, God? A Debate Between a Christian and an Atheist (2004, Craig's opening case includes the moral-grounding deployment); Craig & Moreland, Philosophical Foundations for a Christian Worldview (2003, ch. 24 on ethics); Robert Adams, Finite and Infinite Goods (1999, modified divine command theory); Erik Wielenberg, Robust Ethics (2014, the leading non-theistic alternative, engage it directly).
• Aphorism: "If the evidential POE works, naturalism doesn't. If naturalism does, the evidential POE doesn't have the moral standards it needs."

Tactical notes

• The grounding pin is Craig's signature move on the POE. Internalize his deployment: when the atheist presses the evidential POE, ask "what makes that evil, and what makes you think the standard by which you judge it applies to God?" The atheist who can't ground the standard has no argument; the atheist who grounds the standard via theistic-style moral metaphysics has conceded the framework.
• Don't deploy the grounding pin alone. Pair it with skeptical theism (D1) and the positive theodicy composite (D2). The pin's function is to challenge the atheist's standing to deploy the argument; it does not by itself defeat the argument's content.
• The Lewis "puddles of water" line is rhetorically powerful for popular audiences. Use it.

D4, Natural-evil sub-defense

The empirically-loaded cases for the evidential POE are natural evil, disease, predation, geological disaster, animal suffering across millions of years of pre-human evolution. The free-will defense (which handles moral evil) does not directly address these. Three complementary moves:

D4a, Natural-law defense (Swinburne, Reichenbach)

A world with stable, consistent natural laws is necessary for (a) free moral agency (predictable consequences make moral choice meaningful), (b) scientific reasoning (lawful regularity makes induction possible), and (c) moral responsibility (predictable harm makes harm-avoidance meaningful). A world with stable natural laws inevitably contains some natural evil as collateral: plate tectonics → earthquakes; predation → animal suffering; thermodynamics → entropy and death.

Key insight: God could prevent every specific instance of natural evil only by introducing constant miraculous intervention, which would (1) destroy the lawful regularity necessary for moral agency and science, and (2) make moral character-formation impossible (heroic action requires real danger).

D4b, Augustinian-Fall cosmology (Romans 8:20-22)

The Christian tradition holds that natural evil entered the cosmos consequent on the Fall, the cosmic disorder following human sin. Romans 8.19-22: "the whole creation has been groaning together in the pains of childbirth until now... the creation itself will be liberated." This is a substantive theological position that locates natural evil within the same theodicy as moral evil, both are downstream consequences of the prior moral failure.

Limitations: requires substantive theological commitments (the Fall as a real-historical event with cosmic consequences). Engages with Adam and Eve Historicity and Genesis Hermeneutics. The position is consistent across YEC / OEC / Theistic Evolution but requires the cosmic-consequence claim, which some theistic evolutionists reject.

D4c, Demonic-agency defense (Plantinga, Reichenbach)

Some natural evil may result from fallen-angelic agency. This is a possibility-claim, not an actuality-claim: it shows that natural evil is not in principle inexplicable on theism. Critics call this implausible; defenders point out that implausibility is irrelevant to the modal-possibility bar (this is the same move as Plantinga's transworld depravity for moral evil).

The composite move

The three sub-defenses are complementary. The natural-law defense is the strongest live-debate move (most empirically grounded and apologetically accessible); the Augustinian-Fall is the strongest theologically (most biblically-anchored); the demonic-agency is the strongest modally (lowest bar, possibility, not actuality). Together they cover the space of natural-evil candidate explanations.

Anticipated objections to D4

1. "3.5 billion years of animal suffering before humans existed can't be Fall-consequent." The Augustinian-Fall response struggles with pre-human animal suffering on standard old-earth chronology.

2. "The natural-law defense leaves animal suffering still seemingly gratuitous, animals can't form characters."

3. "Demonic agency is special pleading."

Rebuttals

1. The Augustinian-Fall can be deployed across multiple chronological frameworks. YEC theology resolves this trivially (no pre-Fall animal suffering on a 6000-year chronology). Old-earth Christian theology has standing responses: prolepsis (the Fall's effects propagate back through cosmic time per God's foreknowledge, C.S. Lewis); fallen-angelic-Fall preceding the human Fall (Plantinga, Lewis); the "very good" of Genesis 1:31 was teleologically good, not "no death" good. The Augustinian-Fall is not refuted by old-earth chronology; it requires careful articulation. Failure mode: assuming YEC vs old-earth is settled in favor of old-earth.

2. Animal suffering is engaged by the natural-law defense via the lawful-regularity bind. The Christian needn't claim animal suffering "builds character" in the moral-agent sense, only that it is part of the lawful natural order whose alternative is constant miraculous intervention destroying the conditions for moral agency. Animal sentience and suffering can also be addressed by lesser-sentience arguments (Descartes, extreme; more moderate: Murray's Nature Red in Tooth and Claw, 2008, on neo-Cartesian and adaptive-suffering responses) and eschatological hopes (Isaiah 11:6-9, Revelation 22, the redeemed-creation tradition).

3. The demonic-agency defense is not special pleading because it requires no special evidential bar, it requires only modal possibility. Plantinga's framework here parallels his transworld-depravity move: the atheist must show fallen-angelic agency is necessarily false, not merely implausible. No atheist has done so. Critics conflate "I find this implausible" with "this is logically impossible", the conflation is the failure mode.

Live-cite kit

• Scripture: Romans 8.19-22 (creation groans, will be liberated); Isaiah 11:6-9 (eschatological reconciliation of predator-prey); Revelation 22:1-3 (tree of life leaves for healing of the nations); Job 38-41 (creation theology including animal kingdom).
• Scholarly: Richard Swinburne, Providence and the Problem of Evil (1998, ch. 11, natural-law defense); Bruce Reichenbach, Evil and a Good God (1982, natural-law); Alvin Plantinga, God, Freedom, and Evil (1974, demonic-agency); Michael Murray, Nature Red in Tooth and Claw: Theism and the Problem of Animal Suffering (2008, animal-suffering specifically); Christopher Southgate, The Groaning of Creation (2008, evolutionary theodicy).
• Aphorism: "A lawful world is the price of free agency, and natural evil is the price of a lawful world."

Tactical notes

• The natural-law defense is the most accessible live-debate move on natural evil. Lead with it.
• Don't lead with the demonic-agency defense in popular debate, it sounds like special pleading without academic context. Reserve it for follow-up if the atheist presses the modal-possibility frame.
• Animal suffering is the hardest specific case. Murray's Nature Red in Tooth and Claw is the strongest academic treatment; cite it explicitly in academic engagement.

Master objections to the whole composite

1. "You've just multiplied theodicies endlessly, quantity isn't quality." Reply: the multiplication is over-determination, not desperation. The atheist must defeat ALL of D1, D2, D3, and D4 simultaneously to defeat the composite. Defeating one layer leaves the others intact. This is the same modally-over-determined structure that breaks the logical POE (see Problem of Evil, Free Will Defense), over-determination is a strength, not a weakness. Failure mode: confusing over-determination with desperation.

2. "The Christian has no positive evidence for any of these theodicies." Reply: the dialectic is defensive, not constructive. The atheist is the one asserting the evidence cuts against theism; the theist's defensive work is to show the asserted evidence does NOT cut against theism, by supplying candidate reasons that haven't been ruled out. The theist need not show the candidate reasons are actual, only that they are coherent and unrefuted. (See Plantinga apparatus in Problem of Evil, Free Will Defense for the modal-possibility framing.)

3. "This is all just sophisticated 'God works in mysterious ways.'" Reply: no, skeptical theism is sophisticated epistemic modesty (with the Job-paradigm biblical anchor); the positive theodicy composite positively names candidate reasons (Christological entry, eschatological compensation, soul-making, natural-law collateral). The "mysterious ways" charge is a refusal to engage the actual structure of the response. Failure mode: dismissing a sophisticated apparatus as folk-religious evasion.

4. "This makes the Christian's POE response immune to evidence." Reply: not so, the Christian POE response is engaged by evidence: empirical evidence about the kinds and amounts of suffering (D4 natural-evil work); historical-textual evidence about the Incarnation (D2a Christological work); philosophical evidence about moral grounding (D3 work); philosophical evidence about cognitive asymmetries (D1 work). The atheist who claims the response is unfalsifiable is conflating over-determined with unfalsifiable. Failure mode: confusing over-determination with unfalsifiability.

5. "Rowe's argument was probabilistic; you're defending logical possibility." Reply: the defenses do address probabilistic ground. Skeptical theism (D1) explicitly engages the Bayesian apparatus (Bergmann 2006). The positive-theodicy composite (D2) raises the probability of theism conditional on observed suffering by supplying explanatory candidates. The grounding pin (D3) attacks the inferential probability of the atheist's premise. The natural-evil sub-defense (D4) addresses the empirical anchor cases. All four operate at the appropriate probabilistic level. Failure mode: thinking probabilistic arguments need probabilistic responses in a single mode.

Tactical opening / closing

Opening line: "Let me ask first, are you arguing that God + evil is logically impossible, or that the actual world's evil makes God improbable? Because Mackie conceded in 1982 that the logical version had been answered; the modern argument is the evidential version. If we're discussing the evidential, the response has four layers: skeptical theism on the epistemic side (the inference from 'no apparent reason' to 'no actual reason' fails Wykstra's CORNEA), a positive theodicy composite (Christological + eschatological + soul-making), a grounding question for the moral premise the argument depends on, and a natural-evil sub-defense for the most empirically loaded cases. Which would you like to engage first?"

Force-commit move (when opponent waffles): "Either you can positively identify that no God-justifying reason exists for case E (which requires the cognitive access skeptical theism contests), or you're inferring from 'I see no reason' to 'no reason exists' (which fails CORNEA), or you're claiming this evil case has no candidate-justifying-reason that any theodicy supplies (which requires defeating Christological + eschatological + soul-making simultaneously). Which path are you on?"

Closing landing strip: "The evidential POE has been the live battlefield for 50 years, since Rowe's 1979 paper. The Christian response is well-developed across four layers, epistemic, metaphysical, meta-ethical, and cosmological. Each layer has been field-tested in debate; each has academic-philosophical literature behind it. The argument doesn't break theism, it sharpened theistic response. If you have a specific case you'd like to engage, name it and let's work through the layers."

Live-cite kit (consolidated)

• Scripture (anchor passages): Job 38-41 (the Job-paradigm, God's response to suffering is asymmetry-displaying, not justification-supplying); Romans 8.19-22 (creation groans → liberation); Romans 11:33-36 (depth of God's wisdom); 2 Corinthians 4.17 (light affliction → eternal weight of glory); Revelation 21.4 (every tear wiped); 1 Peter 1:6-7 (trials prove faith).
• Scholarly, skeptical theism (D1): Wykstra 1984 (CORNEA); Alston 1991 (cognitive-asymmetry); Bergmann 2006 (probability-theoretic); see Skeptical Theism.
• Scholarly, Christological theodicy (D2a): Marilyn Adams 1999 (Horrendous Evils); Eleonore Stump 2010 (Wandering in Darkness); William Lane Craig vs Bart Ehrman 2006 debate (live deployment).
• Scholarly, eschatological theodicy (D2b): Craig (Reasonable Faith 2008); Plantinga (Where the Conflict Really Lies 2011, ch. 5).
• Scholarly, soul-making theodicy (D2c): Hick 1966 (Evil and the God of Love); Stump 2010 (the narrative-deepening variant).
• Scholarly, moral grounding (D3): Craig vs Sinnott-Armstrong 2004 (God? A Debate); Craig & Moreland 2003 (Philosophical Foundations, ch. 24); Robert Adams 1999 (Finite and Infinite Goods); engage Wielenberg 2014 (Robust Ethics) for the non-theistic-realist alternative.
• Scholarly, natural-evil sub-defense (D4): Swinburne 1998 (Providence and the Problem of Evil); Reichenbach 1982 (Evil and a Good God); Murray 2008 (Nature Red in Tooth and Claw); Southgate 2008 (The Groaning of Creation).
• Aphorism (consolidated): "Skeptical theism blocks the inference; theodicy supplies the candidates; the grounding pin checks the atheist's standing; the natural-evil sub-defense handles the empirical anchor. Four layers, and the atheist has to beat all four."

Connection to codex

• Master defense: Problem of Evil, Free Will Defense (the logical POE response, Plantinga's apparatus).
• Adjacent / Angstreich-specific: All-Do-Good World POE Defeater (handles Angstreich's modal-polished logical-POE deployment); Jack Angstreich (the deployer, note: not a public deployer of the evidential POE, but this defeater is the standing prep for any opponent who shifts to evidential ground after losing the logical-POE engagement).
• Master concept hub: Problem of Evil.
• D1 anchor: Skeptical Theism (the load-bearing concept hub).
• D3 anchor: Moral Argument (the positive-case version of the grounding pin); Atheism Moral Neutrality Failure (the dedicated meta-ethical reductio).
• Natural-evil neighborhood: Adam and Eve Historicity (the Augustinian-Fall anchor); Genesis Hermeneutics (the cosmic-Fall theology).
• Eschatology: Eschatology synthesis (the over-arching framework for D2b).
• Adjacent atheist objections: Divine Hiddenness / Divine Hiddenness Objection Defeater (the same skeptical-theism move applies; companion defeater).
• Source, debate cite: Jack Angstreich vs Moshi - Logical Possibility All-Do-Good Debate (Tom Rabbittt 2026) (the primary Angstreich corpus, which sticks to logical POE); God and Suffering (existing source hub).

See also

• Problem of Evil, Free Will Defense, master logical-POE defense
• All-Do-Good World POE Defeater, Angstreich-specific logical-POE counter
• Problem of Evil, master concept hub
• Skeptical Theism, D1 concept hub
• Moral Argument, D3 positive-case anchor
• Atheism Moral Neutrality Failure, companion meta-ethical reductio
• Divine Hiddenness Objection Defeater, companion (shared D1 skeptical-theism deployment)
• Free Will Argument from Love, positive case for libertarian freedom
• Eschatology, synthesis for D2b framework
• Adam and Eve Historicity, D4 Augustinian-Fall anchor
• Jack Angstreich, the corpus-specific apologetic context
• William Lane Craig, primary debater-source for D2/D3 deployments
• Bart Ehrman, debater whose Craig-debate motivated much of the deployment shape
• Alvin Plantinga, primary academic source across multiple layers
• Arguments, master index

Common questions this page answers

Q: Why does evil exist if God is good?

The evidential problem of evil rests on a no-see-um inference (I don't see why God would allow this, therefore there's no good reason) that finite cognition cannot reliably make against an infinite mind, and even granting some evils, the Christian framework supplies soul-making goods, free-will-defense, the cross as God's own entrance into suffering, and the eschatological reversal.