# All-Do-Good World POE Defeater ## Intro *"Heaven proves it can be done. People there are free and they never sin. So why did not God just make a world like that to start with?"* This is the modern, philosophically sharpened version of the old "if God is good, why is there evil" objection. The reply rests on a small distinction with big effects. There are two senses of "make something happen." God can directly cause anything that does not require a free creature, like the laws of physics. But when free creatures are involved, He cannot fully and directly cause their free choices, because the choice has to actually come from them, not from Him. The technical names for the two senses are *strong* actualization and *weak* actualization. God strongly actualizes the setting; the creature weakly co-actualizes the free choice inside it. This means a "heaven now" world is logically possible in the abstract, but God cannot guarantee it as a starting condition. He cannot reach in and make sure every free creature freely chooses the good every time without overriding the freedom that makes it count as love at all. He can offer it, set the conditions, and grow people into it, but He cannot strong-arm them there. So heaven is not a counterexample to the free will defense. It is the *outcome* of the defense. It is the world that emerges after creatures have, through trial, redemption, and grace, become the kind of beings who freely will only the good. God did not skip the road because the road is part of how those beings come into existence. Quick reply line: *"God can set the stage. He cannot force a free choice and have it still count as free. Heaven is where the road ends, not where it can start."* ## Cheatsheet **The 30-second reply:** > Heaven is where the road ends, not where it can start. There are two ways God brings things about: He can directly cause what needs no free creature, or He can set the conditions and let free creatures choose inside them. A free choice cannot be directly caused by another agent and still be free, so "making someone freely choose good" is nonsense in the same class as a four-sided triangle. God can offer heaven and grow people into it; He cannot strong-arm free creatures there from day one. **The fast facts:** 1. **Strong vs weak actualization.** God strongly actualizes the setting; the creature weakly co-actualizes the free choice. Free-creaturely worlds can only be weakly actualized, even by omnipotence. 2. **Possibility, not actuality.** Plantinga does not claim transworld depravity is true, only that it is coherently possible; a possibility is all it takes to defeat an impossibility claim. 3. **Mackie conceded.** J. L. Mackie, architect of the modern logical problem of evil, granted in The Miracle of Theism (1982) that Plantinga's defense succeeded. 4. **Heaven is consummated, not initial.** Aquinas's non posse peccare (inability to sin) is path-dependent on prior earthly formation, a different modal state from initial freedom. 5. **Classical omnipotence is restricted.** Aquinas, Anselm, and Augustine all define it as power over the doable, not power over contradictions. **The 3 strongest counter-moves:** - *"Logical or evidential?"* Force the disambiguation: if the logical problem of evil, Mackie already conceded it in 1982; if evidential, that is a different debate for skeptical theism and soul-making. - *"Can God make someone freely choose?"* If yes, the choice was not free; if no, then A1's unrestricted omnipotence requires logical-grammatical nonsense. - *"Heaven equivocates."* Agree heaven exists, then refuse the leap from consummated freedom to initial freedom; the destination cannot be the starting point. **Reciprocal concessions (grant the small point, then collect a bigger one):** - *Grant:* An all-do-good world is genuinely logically possible in the abstract. *Now collect:* Logical possibility is not feasibility for God; a world can be possible while no world God can weakly actualize contains it, which is exactly where the defense lives. - *Grant:* Heaven really is a world of free creatures who never sin. *Now collect:* Then admit heaven is the consummated end of a formation path, so it proves a sin-free destination is possible, not a sin-free starting condition. - *Grant:* God's power is in a sense conditional on what free creatures would do. *Now collect:* That conditional restriction is the classical doctrine of omnipotence, so your objection only works by importing a voluntarist folk concept the tradition already rejected. **The closing line:** > The free will defense was only ever the answer to the impossibility charge, and that charge has been dead since 1974, by Mackie's own admission. ## In full A modal-polished version of Mackie's 1955 logical Problem of Evil has become the central deployment of philosophically-literate atheist debaters (paradigmatically [Jack Angstreich](/codex/jack-angstreich/) in his Moshi-debate corpus). The argument runs: *"God's omnipotence is the capacity to actualize any logically possible state of affairs; a world in which significantly free creatures always freely choose the good is a logically possible state of affairs (heaven is the proof of concept); therefore God could have actualized that world, and his failure to do so is incompatible with his omnibenevolence."* This page is structured as debate prep for live engagement with that specific deployment. It does **not** replace [Problem of Evil, Free Will Defense](/codex/problem-of-evil-free-will-defense/), which contains the master defense, but rather supplies per-premise rebuttals tailored to the atheist's polished modal form, the brute-contingency pin against libertarian free will, and the compatibilism trap. The single most important live-debate move is **distinguishing strong actualization from weak actualization**: God's omnipotence does not include the power to *strongly* actualize states of affairs containing creaturely free choices, because such states are weakly actualized via counterfactuals of creaturely freedom (CCFs) which God does not directly control. Once that distinction lands, the argument collapses. Companion to [Problem of Evil, Free Will Defense](/codex/problem-of-evil-free-will-defense/) (master), [Skeptical Theism](/codex/skeptical-theism/) (evidential-POE companion), and [Free Will Argument from Love](/codex/free-will-argument-from-love/) (libertarian-freedom positive case). ## Argument structure **The atheist's argument (Angstreich's modal-polished form of Mackie):** | # | Premise | |---|---| | **A1** | God's omnipotence = capacity to actualize any logically possible state of affairs. | | **A2** | A world in which significantly free creatures always freely choose the good is a logically possible state of affairs (heaven is the proof of concept). | | **A3** | If God could have actualized such a world and is omnibenevolent, He would have. | | **AC** | **Therefore the actual world's containing evil is incompatible with God's existence as classically conceived.** | **The Christian defense:** | # | Premise | |---|---| | **D1** | Omnipotence is the capacity to **strongly** actualize any logically possible *intrinsic* state of affairs, not the capacity to strongly actualize states of affairs whose obtaining requires creaturely *free* choices. | | **D2** | Worlds containing significantly free creatures who always freely choose the good are *weakly* actualizable only, God can set up the circumstances, but the obtaining of the world's good-content depends on counterfactuals of creaturely freedom (CCFs) that God does not control. | | **D3** | It is *possible* that for every significantly-free creaturely essence, there is at least one circumstance in which that essence would freely choose evil (Plantinga's **transworld depravity**, *coherent* as a modal possibility, *not* asserted as actual). | | **D4** | Therefore it is *possible* that *no* feasible world of significantly-free creatures is sin-free, even with God omnipotent and omnibenevolent. | | **DC** | **The argument's A1-A2-A3 chain does not entail AC. The all-do-good world's logical possibility does not entail its feasibility for God.** | ## Form **Defensive / possibility-based (modal); reductio against opponent's equivocation on "actualize."** The defense does not claim transworld depravity is *actually* true, only that it is *possibly* true. Modal possibility suffices to defeat the logical-impossibility claim Mackie originally made and Angstreich modernizes. The defense exposes a critical equivocation in A1: it conflates *strong* actualization (God directly brings about) with the generic capacity-to-actualize (which for free-creaturely worlds is *weak*, God sets up circumstances, creatures freely choose). Mackie himself conceded the defense's success in *The Miracle of Theism* (1982); Angstreich's modal polish does not revive the logical POE, it merely restates it more carefully and is defeated by the same Plantinga apparatus. --- ## D1, Omnipotence: strong vs weak actualization ### Affirmative case (second-order arguments) 1. **The classical doctrine of omnipotence has always required logical coherence.** Aquinas (*ST* I q. 25 a. 3): God's omnipotence is the capacity to do whatever does not involve contradiction. Anselm, Augustine, and the entire scholastic tradition concur. The atheist who frames omnipotence as bare "any logically possible state of affairs" needs to acknowledge that the *coherence* of the state of affairs given the *kinds of entities involved* is part of the logical-possibility analysis. A world in which God *strongly* actualizes a free creaturely choice is incoherent in the same way as a four-sided triangle: the "free" part rules out the "strongly-actualized-by-another-agent" part. 2. **Plantinga's strong-vs-weak distinction is foundational, not novel.** Plantinga (*Nature of Necessity*, 1974, §9.5) gives the canonical formulation: *God strongly actualizes S* = God brings about S directly; *God weakly actualizes S* = God strongly actualizes some state of affairs S* such that S obtains as a consequence (via creaturely free choices God did not directly cause). Worlds containing creaturely free actions can only be weakly actualized, even by an omnipotent God, because the free-action component is constitutively not God's doing. 3. **The distinction is intuitive once stated.** If God can *make* you freely choose chocolate, then your choice was not free; if it was free, then God did not *make* you choose it. "Making someone freely choose X" is logical-grammatical nonsense ("forced free choice"). The atheist's A1, applied to free-creaturely worlds, requires logical-grammatical nonsense. The defense exposes the nonsense; the atheist must either retract A1 or commit to nonsense. ### Anticipated objections 1. **"This is just verbal trickery, God could still arrange the circumstances so the free choices all go right."** The atheist refines: God doesn't have to *force* the choices, He just has to set up the world so the choices happen to all go right. 2. **"You're stipulating that 'omnipotence' has a restricted meaning that conveniently saves theism."** The definitional-rigging charge. 3. **"What about Frankfurt-style examples, God could be a backup mechanism that only intervenes if the creature is about to choose evil?"** A more sophisticated objection from contemporary action theory. ### Rebuttals 1. **"Arranging circumstances so all choices go right" is precisely *weak* actualization, and it requires controlling the counterfactuals of creaturely freedom, which is the contested move.** Once the atheist concedes that the divine action is *circumstance-setting* rather than choice-causing, they have conceded the strong-vs-weak distinction. The remaining question, "can God set up circumstances such that all free choices in fact go right?", depends on what the CCFs are. Plantinga's transworld depravity hypothesis (D3 below) is precisely the *modal possibility* that for every creaturely essence, no circumstance-set yields uniformly-good free choices. **The atheist's refined objection is no longer a logical objection, it is an empirical / probabilistic claim about what CCFs in fact obtain.** That shifts to the evidential POE, where different rebuttals (skeptical theism, soul-making, Christological theodicy) take over. **Failure mode: confusing logical with evidential POE.** 2. **The restricted meaning is the *historical* meaning, not a recent dodge.** Aquinas, Anselm, Augustine, Maimonides, and Avicenna all defined omnipotence as power-over-what-is-doable. The classical theist position has never required God to be capable of contradictions. The atheist who insists on the unrestricted reading is appealing to a *folk* concept of omnipotence ("can God do *anything*?") that no major theistic philosopher has ever held. The folk concept is a strawman. **Failure mode: imposing folk omnipotence on classical theism.** 3. **Frankfurt-style "backup-mechanism" worlds collapse into either compatibilism or non-libertarian freedom.** In a genuine Frankfurt case, the agent's actual choice is *not* causally affected by the backup mechanism, but the *possibility* of choosing otherwise is removed. Philosophers of action are divided on whether such cases preserve libertarian responsibility (Frankfurt's original claim) or whether they actually require compatibilism (the Widerker / Kane objection). The defense doesn't need to resolve this, it needs only to note that Frankfurt-style proposals are *contested in the action-theory literature*, not a clean knockdown. If the atheist deploys Frankfurt cases, they have entered a technical action-theory dispute, not a logical-POE refutation. **Failure mode: outsourcing the argument to a contested action-theory thesis.** ### Live-cite kit - **Scripture**: [James 1:13](/codex/james-1-13/) ("God tempts no one", God does not actively bring about evil choices); [Genesis 1.27](/codex/genesis-1-27/) (*imago Dei* = significantly-free creaturely capacity); [Deuteronomy 30.19](/codex/deuteronomy-30-19/) ("I have set before you life and death... choose life", God presents the choice; the creature chooses). - **Scholarly**: Plantinga (*Nature of Necessity*, 1974, §9.5, strong vs weak actualization); Plantinga (*God, Freedom, and Evil*, 1974, ch. I §9, defense form); Aquinas (*ST* I q. 25 a. 3, omnipotence restricted to non-contradictory action); Flint (*Divine Providence: The Molinist Account*, 1998, ch. 2-3, weak actualization formalized). - **Aphorism**: "*Making* someone *freely* choose X is in the same impossibility class as a four-sided triangle." ### Tactical notes - **Lead with the strong-vs-weak distinction immediately.** Most popular-level apologetic engagements never reach this point because the Christian apologist tries to defend A1 as stated. The right move is to refuse A1 as a careless characterization of omnipotence, and supply the more careful version (D1). This re-roots the entire argument in the Plantinga apparatus, where the existing defense applies. - **The "make someone freely choose" line is the killer aphorism.** Use it. The audience hears the absurdity. - **What NOT to do live**: defend the bare logical possibility of an all-do-good world. The Christian theologian *grants* that possibility (Plantinga grants it). The defense is at the *feasibility* step, not the *possibility* step. Granting possibility while denying feasibility is the dialectical heart of the defense; conflating them concedes the argument. --- ## D2, Free-creaturely worlds are weakly actualizable only ### Affirmative case (second-order arguments) 1. **The metaphysics of free choice precludes strong actualization.** A "significantly free" choice is, by definition, a choice not necessitated by causal antecedents external to the agent. If God strongly actualizes the choice, it is necessitated by an external causal antecedent (God's volition), and is therefore not free. The two predicates ("free" and "strongly actualized by God") are jointly inconsistent. This is not a contingent restriction on God's power; it is a logical-grammatical restriction on the meaningful application of the concepts. 2. **Counterfactuals of creaturely freedom (CCFs) carry the metaphysical load.** Under Molinism and similar middle-knowledge accounts (Luis de Molina 1588; Flint 1998; Craig 1987), God knows for each creaturely essence the counterfactual: *if creature C were placed in circumstance S, C would freely do X*. God's choice of which circumstances to instantiate is the mechanism of weak actualization, God controls *which* free-creaturely actions get realized by choosing which circumstances obtain. But God does *not* control the CCFs themselves, those are facts about what the creature would freely do, grounded in the creaturely essence. 3. **Open Theism over-determines the defense.** If counterfactuals of creaturely freedom are incoherent (William Hasker 1989; some Reformed critics), then the Open Theist alternative blocks the all-do-good objection differently: God doesn't know in advance how free creatures will choose, because there is no future-tensed fact to know. God permits the world of free creatures and meets it as it unfolds. The all-do-good world is *not* a world God could foresee and actualize from creation; it would require continuous corrective intervention that violates the freedom-condition. Either way (Molinist CCFs or Open Theist not-yet-fact), strong actualization of free-creaturely worlds is excluded. ### Anticipated objections 1. **"The grounding objection to Molinism: what makes CCFs true?"** Jack Angstreich explicitly names this. If CCFs have no truthmaker, they're not true, and the Molinist machinery collapses. 2. **"Even granting CCFs, God could have chosen to instantiate only the essences with all-good CCFs."** If God knows the CCFs, He can pre-select. 3. **"Heaven is the existence proof that an all-do-good world is feasible, so transworld depravity is empirically refuted."** Jack's deployment of the heaven counterexample. ### Rebuttals 1. **The grounding objection is contested but not decisive, and the defense is over-determined.** Multiple grounding accounts have been proposed: creaturely-essence-properties as truthmakers (Flint *Divine Providence* ch. 5; Freddoso); divine-omniscience-of-modal-truths (Plantinga); subjunctive-conditional-primitivism (Adams). The grounding objection is a *live* philosophical dispute, not a closed case. Even if it succeeded, the defense has Open Theism as backup (no CCFs needed, God simply doesn't know future free choices in advance). **The atheist must defeat both the Molinist defense *and* the Open Theist defense to defeat the Free Will Defense. No atheist has done so.** **Failure mode: targeting one Christian metaphysical package while ignoring the alternatives.** 2. **The "pre-select essences with all-good CCFs" objection assumes there *are* such essences, which is precisely what transworld depravity denies.** Plantinga's hypothesis is that it is *possible* that for *every* significantly-free creaturely essence, at least one circumstance yields a free evil choice. If TWD is possible (it has not been refuted), then no selection of essences yields a sin-free world. The objection presupposes the defense is wrong; it does not argue the defense is wrong. **Failure mode: question-begging.** 3. **Heaven is *not* the existence proof Angstreich claims, heaven is path-dependent on prior creaturely formation.** Christian theology has always held that heaven is the *consummated* state of moral formation, what Aquinas (*ST* I-II q. 5 a. 4) calls *non posse peccare* (the inability to sin, consequent on beatific vision) rather than *posse non peccare* (the ability not to sin, characteristic of pre-fall humanity). Heaven's inhabitants are confirmed-in-good through prior earthly trial; they are not creatures who started in heaven. Augustine (*De Civitate Dei* XXII.30), Aquinas, and the entire scholastic tradition are uniform on this. **Heaven proves that a sin-free *consummated* state of free creatures is possible, it does *not* prove that a sin-free *initial* state of free creatures is possible. These are different modal claims.** Angstreich's heaven counterexample equivocates between them. **Failure mode: equating consummated freedom with initial freedom.** ### Live-cite kit - **Scripture**: [Romans 8:28-30](/codex/romans-8-28-30/) (predestination as God's foreknown response to free agents; God works through, not around, creaturely freedom); [1 Corinthians 13:9-12](/codex/1-corinthians-13-9-12/) ("now we see in a mirror dimly... then we shall see face to face", eschatological consummation as the climactic moral-epistemic state, not the initial); [Revelation 21:4-5](/codex/revelation-21-4-5/) ("behold I make all things new", heaven as renewed-consummated, not initial-created). - **Scholarly**: Plantinga (*God, Freedom, and Evil* §I.9); Flint (*Divine Providence: The Molinist Account*, 1998, esp. ch. 5 on grounding); Freddoso (intro to translation of *On Divine Foreknowledge*, 1988); Hasker (*God, Time, and Knowledge*, 1989, Open Theism); Aquinas (*ST* I-II q. 5 a. 4 on beatific vision and *non posse peccare*); Augustine (*De Civitate Dei* XXII.30). - **Aphorism**: "Heaven is the destination, not the starting point, the all-do-good world *is* the destination." ### Tactical notes - **The heaven-equivocation reply is the most important single tactical move on this premise.** Most apologetic engagements lose at this step because the apologist agrees with Angstreich's heaven-counterexample framing. The right move is to *agree heaven exists* while *refusing the inferential leap* from consummated-freedom to initial-freedom. The path-dependence framing (Aquinas; Hick's soul-making) does the work. - **Don't get bogged down in the grounding-objection literature live.** It is a deep and active dispute, citing Flint or Freddoso is enough to show the defense has substance. Pursuing the technicalities will lose the audience. - **Force-commit move**: "Are you arguing the *logical* POE (impossible for God + evil) or the *evidential* POE (improbable for God given the actual evil)? Because Plantinga's defense answered the logical POE, Mackie himself conceded it in 1982. If you're arguing the evidential POE, that's a different argument requiring a different defense ([Skeptical Theism](/codex/skeptical-theism/), [Hick Soul-Making Theodicy](/codex/soul-making-theodicy/))." --- ## D3, Transworld depravity is coherently possible ### Affirmative case (second-order arguments) 1. **The defense requires only modal possibility, not actuality.** Plantinga does *not* argue that transworld depravity is *true*. He argues that it is *coherent*, that the proposition "for every significantly-free creaturely essence, there is at least one circumstance in which that essence would freely sin" is not self-contradictory. Coherence is sufficient to defeat the logical-impossibility claim of Mackie / Angstreich. 2. **Empirical fallenness is consistent with TWD.** Every observed society of free creatures has contained moral evil. This does not prove TWD; it does mean the hypothesis is empirically open. [Romans 3:23](/codex/romans-3-23/) ("all have sinned and fall short of the glory of God") is the textual locus of the Christian's empirical claim that universal moral fallenness obtains in the actual world. 3. **The defense is over-determined modally.** Even if TWD is false in some possible worlds, Plantinga's defense requires only that it be *possible*, i.e., true in at least one possible world. This is an extraordinarily weak modal claim. Defeating it requires showing TWD is *necessarily false*, which no atheist philosopher has accomplished. ### Anticipated objections 1. **"Transworld depravity is wildly implausible."** Even if logically possible, it strains credulity that *every* creaturely essence would freely sin in *some* circumstance. 2. **"What about creatures God hasn't created yet? Couldn't there be a non-TWD essence among them?"** A modal-existence quantifier challenge. 3. **"This makes God look helpless, bound by the creaturely essences He didn't choose."** The divine-sovereignty objection. ### Rebuttals 1. **Implausibility is not impossibility, and the defense's bar is possibility.** The atheist who retreats to "implausibility" has conceded the logical POE has failed and is now arguing the evidential POE, where different rebuttals apply (skeptical theism: we are not in a cognitive position to judge probabilities over creaturely essences; soul-making: even apparently-gratuitous moral evil can serve character-formation). **Failure mode: confusing implausibility with impossibility.** 2. **Creaturely essences are not God's choice, they are conceptual possibilities God either instantiates or doesn't.** Plantinga's view (and the standard Molinist view) is that creaturely essences are *abstract objects* whose properties are not under God's control. God knows the CCFs for each essence; God chooses which essences to instantiate; God does not choose what the essences are. The objection assumes God *creates* creaturely natures from nothing, which conflates instantiation with constitution. **Failure mode: misunderstanding the metaphysics of creaturely essence.** 3. **The "helpless God" framing imposes a non-classical view of sovereignty.** Classical theism has never held that God's sovereignty entails control over abstract objects (numbers, logical truths, possible worlds, creaturely essences). God's sovereignty is over the *actual*, over what is instantiated. This is the classical doctrine of divine aseity (see [Aseity](/codex/aseity/)) and the eternal-ideas-in-God-as-archetypes view (Augustine, Aquinas). The objection imports a voluntarist conception of divine sovereignty (Descartes, Ockham) that most classical theists have explicitly rejected. **Failure mode: assuming voluntarist sovereignty as the default.** ### Live-cite kit - **Scripture**: [Romans 3.23](/codex/romans-3-23/) ("all have sinned"); [Romans 5.12](/codex/romans-5-12/) ("through one man sin entered the world... and so death spread to all men"); [Genesis 6:5](/codex/genesis-6-5/) ("every intention of the thoughts of his heart was only evil continually"). - **Scholarly**: Plantinga (*Nature of Necessity*, 1974, §9.4, transworld depravity); Plantinga (*God, Freedom, and Evil*, 1974, pp. 44-53); Mackie's concession (*The Miracle of Theism*, 1982, p. 154); Howard-Snyder & O'Leary-Hawthorne ("Transworld Sanctity and Plantinga's Free Will Defense," *International Journal for Philosophy of Religion* 1998, discusses limits). - **Aphorism**: "Plantinga doesn't say transworld depravity is true. He says it's not contradictory, and that's enough." ### Tactical notes - **The modal-possibility framing is the defense's strength.** Restate it explicitly: "I am not claiming transworld depravity is *true*. I am claiming it is *possibly true*, and possibility is enough to defeat the logical-impossibility claim." - **Mackie's concession is the strongest single citation.** "Mackie, the architect of the modern logical POE, conceded in *The Miracle of Theism* that Plantinga's defense succeeded. The dialectic has moved to the evidential POE." This forces the opponent to either dispute Mackie (which they can't, the text is explicit) or to admit the logical POE has been answered. --- ## D4, Therefore the all-do-good world's logical possibility does not entail its feasibility for God ### Affirmative case (second-order arguments) 1. **The conjunction D1-D2-D3 yields D4 by direct modal-logical inference.** If omnipotence is restricted to non-contradictory action (D1), and free-creaturely worlds are weakly actualizable only (D2), and transworld depravity is a coherent modal possibility (D3), then it follows by elementary modal logic that the all-do-good world may be *logically possible* (some world contains it) while being *infeasible for God* (no world *God can weakly actualize* contains it). The atheist's A1-A2-A3 chain ignores this distinction. 2. **Feasibility, not bare logical possibility, is the relevant modal predicate.** Plantinga's terminology: a possible world W is *feasible for God* iff the CCFs that obtain are such that God can weakly actualize W. Some possible worlds are not feasible for God (those whose CCFs God cannot bring about by setting circumstances). The all-do-good world is *possibly* not feasible, and that possibility is sufficient to break A1-A2-A3-AC. 3. **The defense's over-determination via Open Theism reinforces the conclusion.** Even if Molinism is false (CCFs are incoherent), Open Theism gives a different reason for the same conclusion: God doesn't know future free choices in advance, so couldn't pre-construct an all-do-good world. The conclusion D4 holds under both Christian metaphysical packages. ### Anticipated objections 1. **"You've just relabeled 'possible-for-God' as 'feasible', that's a verbal trick."** The terminological-rigging charge. 2. **"This makes God's omnipotence look weirdly conditional, it depends on what creatures would do, which God doesn't control."** The substantive-doctrine objection. 3. **"Even granted feasibility, why didn't God just not create the depraved essences?"** The selection-objection revival. ### Rebuttals 1. **The strong-vs-weak distinction is *not* relabeling, it is the analysis of what 'possible-for-God' actually means for states-of-affairs-containing-free-choices.** The atheist who calls this verbal trickery has not engaged the metaphysics: the question *what does it mean to actualize a state of affairs containing a free choice* requires an answer, and the strong-vs-weak distinction is the answer. If the atheist proposes an alternative analysis, let them give it, but they cannot wave the question off as terminological. **Failure mode: dismissing metaphysics as semantics.** 2. **Conditional omnipotence is the *classical* view.** Aquinas (*ST* I q. 25 a. 3) explicitly restricts omnipotence to the non-contradictory; the conditional structure (God can do anything *that is doable given the kinds of entities involved*) is classical theism's default. The atheist's "weird" objection presupposes the voluntarist conception of omnipotence (Descartes, Ockham) as the default, but that is itself a minority view in the Christian tradition. **Failure mode: imposing voluntarist omnipotence on classical theism.** 3. **The "selection" objection assumes the existence of non-depraved essences, which TWD denies as a *modal possibility*.** If for every significantly-free essence there is at least one circumstance yielding free evil choice, then no selection of essences gives a sin-free world. The objection assumes there *are* selectable non-depraved essences; the defense's possibility-claim denies this. The objection begs the question. **Failure mode: presupposing the defense's possibility-claim is false.** ### Live-cite kit - **Scripture**: [2 Timothy 2:13](/codex/2-timothy-2-13/) ("if we are faithless, He remains faithful, for He cannot deny Himself"), God's omnipotence is restricted by God's nature, classical scope; [Numbers 23:19](/codex/numbers-23-19/) (God's faithfulness as constitutive). - **Scholarly**: Plantinga (*God, Freedom, and Evil*, 1974, the consolidated argument); Flint (*Divine Providence: The Molinist Account*, 1998, the modal apparatus); Mackie (*The Miracle of Theism*, 1982, concession at ch. 9); Aquinas (*ST* I q. 25 a. 3, restricted omnipotence). - **Aphorism**: "Logical possibility says the world could exist; feasibility says God could bring it about. Plantinga's defense lives in the gap." ### Tactical notes - **The summarized conclusion is the close**: "So the argument has been: *yes*, an all-do-good world is logically possible; *no*, that doesn't entail God could have brought it about; God's omnipotence is power over the doable, free creatures introduce the question of feasibility-given-CCFs, and the defense requires only the possibility, not the actuality, of universal transworld depravity. Mackie conceded in 1982 that this defense defeated the logical POE. If you have a substantive disagreement with Mackie's concession, let's hear it. Otherwise, the dialectic has moved to the evidential POE, and that's a different debate." --- ## The brute-contingency pin (Angstreich's secondary move) Angstreich's secondary attack: when the Christian replies "God can't *guarantee* the free choices," Jack pins LFW on either of two horns: | Horn | Charge | Christian reply | |---|---|---| | **Brute contingency** | LFW choices have no causal explanation; "freedom" = randomness | **Agent causation**: choices are grounded in the *agent* (substantial self with intellect-will-character) without being determined by external antecedent conditions. Neither determined nor random. (Chisholm; Goetz; O'Connor; Clarke; Hasker; Lowe.) | | **Compatibilism** | If choices ARE explained (desires + character), God could engineer them | This horn is only relevant to *compatibilist* Christians ([Calvinism](/codex/calvinism/) / Reformed compatibilism); the LFW Christian rejects the antecedent. Compatibilists answer via the moral-formation-through-creaturely-encounter line. | The pin presupposes the dichotomy *random or determined*. Agent causation is a third option, contested in action theory but live. The defense doesn't need to win the action-theory debate, only to show the dichotomy is not exhaustive. ### Suggested deployment - "Your dichotomy assumes choices are either *determined by external antecedents* or *random*. There is a third option: *agent causation*. The agent, the substantial self with intellect, will, and character, is the causal originator of the choice. The choice is grounded in the agent without being determined by anything external to the agent. This is the position of Chisholm, Goetz, O'Connor, and the contemporary libertarian action-theory literature. If you want to claim agent causation is incoherent, that's an argument you'd have to make, and the literature does not regard it as closed." - See [Libertarian Free Will](/codex/libertarian-free-will/) for the dedicated hub. For now, the *Stanford Encyclopedia of Philosophy* entries on "Agent Causation" (Timothy O'Connor) and "Incompatibilism: Arguments for" are useful primary cites. --- ## Master objections to the whole argument 1. **"This is all just Plantinga warmed over, it doesn't address my modal-polished version."** Reply: Plantinga's framework *does* address the modal-polished version, because Plantinga's framework *is* itself a modal account. The strong-vs-weak distinction, transworld depravity, and the over-determination via Molinism + Open Theism are the *core* of the Free Will Defense. Restating Mackie's challenge in cleaner modal language doesn't escape the apparatus that handles Mackie's challenge. 2. **"The grounding objection to Molinism still stands."** Reply: granted that the grounding objection is *live*, it is. But (a) the Molinist literature has direct responses (Flint, Freddoso, Plantinga); (b) Open Theism doesn't need CCFs at all and gives an alternative path to the same defense; (c) the *possibility* of either framework being correct is enough, the defense is modally over-determined. The atheist must defeat *both* the Molinist apparatus *and* the Open Theist alternative simultaneously to defeat the Free Will Defense. No one has done so. 3. **"Heaven proves an all-do-good free world is feasible."** Reply: heaven proves a *consummated* state of free creatures is sin-free. Path-dependence on prior earthly moral formation is the classical view (Aquinas, Augustine, Hick's soul-making). The argument equivocates between consummated and initial freedom. They are different modal states. 4. **"The defense saves God's existence at the cost of making Him less than omnipotent."** Reply: God's omnipotence has always been classical-restricted (power over the doable, not the contradictory). The atheist imposes a voluntarist / folk-omnipotence view on classical theism. Aquinas, Anselm, Augustine all explicitly affirm the restriction; the defense is in the classical mainstream, not a recent dodge. 5. **"Even granted the defense, the actual *amount* of evil is still incompatible with God."** Reply: that is the evidential POE, a different argument requiring different replies. [Skeptical Theism](/codex/skeptical-theism/), [Hick Soul-Making Theodicy](/codex/soul-making-theodicy/), *Greater-Good Theodicies*, the Christological theodicy (cross-incarnational entry into suffering) are the relevant responses. The Free Will Defense was always *only* the answer to the *logical* POE. Granting the move to evidential ground is conceding the defense has done its work. ## Tactical opening / closing **Opening line:** "Before we go further, are you arguing the *logical* problem of evil (it's *impossible* for God to coexist with evil) or the *evidential* problem of evil (it's *improbable* given the actual evil)? Because Mackie, who wrote the original logical-POE paper in 1955, conceded in 1982 that Plantinga's defense defeated his argument. So if you're arguing the logical version, I'll explain why; if you're arguing the evidential, the responses are different. Which is it?" **Force-commit move (when opponent waffles):** "Either an all-do-good world is *logically* impossible for God (which you're not claiming, you grant it's logically possible) or it's *feasibly* impossible for God given the counterfactuals of creaturely freedom (which is exactly what the defense argues). Which alternative are you taking? If neither, then your argument doesn't follow." **Closing landing strip:** "The defense doesn't claim it has the *theodicy*, the full explanation of why God permits each particular evil. It claims the *logical* impossibility charge has failed. That charge has failed since 1974, and Mackie himself acknowledged it. The remaining work is the *evidential* problem of evil, where skeptical theism and the various theodicy traditions take over. The Free Will Defense has done its job: the conjunction *God + evil* is not a contradiction." ## Live-cite kit (consolidated) - **Scripture:** [Genesis 1.27](/codex/genesis-1-27/) (*imago Dei*); [Deuteronomy 30.19](/codex/deuteronomy-30-19/) (choose); [Romans 3:23](/codex/romans-3-23/) (universal sin); [James 1:13](/codex/james-1-13/) (God tempts none); Aquinas's exegesis of God's nature in *ST* I qq. 19-25. - **Scholarly, Plantinga apparatus:** *God, Freedom, and Evil* (1974); *The Nature of Necessity* (1974, §9); *Warranted Christian Belief* (2000). - **Scholarly, Mackie concession:** *The Miracle of Theism* (1982, ch. 9, p. 154). - **Scholarly, Molinist apparatus:** Flint, *Divine Providence: The Molinist Account* (1998); Freddoso, intro to Molina's *On Divine Foreknowledge* (1988); Adams, "Middle Knowledge and the Problem of Evil" (*APQ* 1977). - **Scholarly, Open Theist alternative:** Hasker, *God, Time, and Knowledge* (1989); Pinnock et al., *The Openness of God* (1994). - **Scholarly, agent causation (for the brute-contingency horn):** Chisholm, "Human Freedom and the Self" (1964); O'Connor, *Persons and Causes* (2000); Goetz, *Freedom, Teleology, and Evil* (2008). - **Aphorism**: "*Possibility* defeats *impossibility*. The defense is modest by design, and modesty wins this argument." ## Connection to codex - **Companion**: [Problem of Evil, Free Will Defense](/codex/problem-of-evil-free-will-defense/), the master Free Will Defense syllogism; this defeater is the Angstreich-specific debate-prep deployment. - **Master concept hub**: [Problem of Evil](/codex/problem-of-evil/), situates the defense within the broader theodicy / defense landscape. - **Entity**: [Jack Angstreich](/codex/jack-angstreich/), the deployer; see his entity hub for the full tactical-engagement notes. - **Adjacent**: [Skeptical Theism](/codex/skeptical-theism/) (evidential-POE companion); [Hick Soul-Making Theodicy](/codex/soul-making-theodicy/) (theodicy companion); [Evil as Privation of Good](/codex/evil-as-privation-of-good/) (metaphysics of evil); [Free Will and Determinism](/codex/free-will-and-determinism/) (intra-Christian dispute); [Molinism](/codex/molinism/) (the grounding-objection literature); [Free Will Argument from Love](/codex/free-will-argument-from-love/) (positive case for LFW). - **Further**: J.L. Mackie, architect of the logical POE. ## See also - [Problem of Evil, Free Will Defense](/codex/problem-of-evil-free-will-defense/), master defense - [Problem of Evil](/codex/problem-of-evil/), concept hub - [Jack Angstreich](/codex/jack-angstreich/), atheist debater - [Sye Ten Bruggencate vs Jack Angstreich - Determinism Debate (Godless Girl 2026)](/codex/sye-ten-bruggencate-vs-jack-angstreich-determinism-debate-godless-girl-2026/), companion debate source - [Jack Angstreich vs Moshi - Logical Possibility All-Do-Good Debate (Tom Rabbittt 2026)](/codex/jack-angstreich-vs-moshi-logical-possibility-all-do-good-debate-tom-rabbittt-2026/), primary debate source - [Skeptical Theism](/codex/skeptical-theism/), evidential-POE companion - [Free Will and Determinism](/codex/free-will-and-determinism/), intra-Christian dispute - [Molinism](/codex/molinism/), middle-knowledge framework - [Free Will Argument from Love](/codex/free-will-argument-from-love/), positive case for libertarian freedom - [Arguments](/codex/arguments/), master index
## Common questions this page answers **Q: If heaven has free people who never sin, why did not God just make a world like that to begin with?** Heaven is the consummated outcome of a moral path, not the starting point. Aquinas calls heaven's *non posse peccare* (inability to sin) the confirmed-in-good state that follows the beatific vision; it is path-dependent on prior earthly formation. God cannot start free creatures already at the destination without bypassing the trial that makes the destination possible. **Q: Why can't an all-powerful God make people who freely always choose good?** There are two ways God can bring something about. He can directly cause it (strong actualization), or He can set up the conditions and let it happen (weak actualization). Free choices, by definition, cannot be strongly actualized by another agent; "making someone freely choose X" is logical-grammatical nonsense in the same class as a four-sided triangle. **Q: What is the free will defense?** The defense, developed by [Alvin Plantinga](/codex/alvin-plantinga/) in *God, Freedom, and Evil* (1974), shows that the coexistence of God and evil is not a logical contradiction. It rests on strong-vs-weak actualization plus transworld depravity (the modal possibility that for every significantly-free creaturely essence, at least one circumstance yields a free evil choice). J. L. Mackie himself conceded the defense's success in *The Miracle of Theism* (1982). **Q: Does Plantinga claim transworld depravity is actually true?** No. He claims it is *coherently possible*, not actually true. The logical problem of evil claims God-plus-evil is impossible; defeating an impossibility claim requires only showing a possibility. Plantinga's modal claim is extraordinarily weak by design, which is why it works. **Q: Doesn't this make God less than omnipotent?** No, classical omnipotence has always been restricted to the non-contradictory. Aquinas (*ST* I q. 25 a. 3), Anselm, Augustine, and the entire scholastic tradition define omnipotence as power over what is doable, not power over contradictions. The "weird conditional" objection imports a voluntarist conception of omnipotence that classical theists have explicitly rejected. **Q: Is the problem of evil still a problem after this defense?** The *logical* problem of evil (impossibility) has been answered since 1974. The *evidential* problem of evil (improbability given the amount or kind of evil) remains a live discussion, addressed by [Skeptical Theism](/codex/skeptical-theism/), [Hick Soul-Making Theodicy](/codex/soul-making-theodicy/), and the Christological theodicy. Granting the move to evidential ground is conceding the free will defense has done its work.