You are viewing a free preview of this lesson.
Subscribe to unlock all 10 lessons in this course and every other course on LearningBro.
The free will defence is the most widely respected response to the logical problem of evil (Lesson 1), and one of the three responses named in the AQA 7062 specification (with Hick's soul-making theodicy and Griffin's process theodicy). Its modern form is the work of the American philosopher Alvin Plantinga (b. 1932), principally in God, Freedom, and Evil (1974). A single point of terminology governs everything that follows: Plantinga offers a defence, not a theodicy. A theodicy claims to give God's actual reasons for permitting evil; a defence has the far more modest aim of showing that the existence of God and the existence of evil are logically compatible — that there is no contradiction in affirming both. To do that, Plantinga needs only to identify one proposition that is possibly true and that, if true, would reconcile the two. He does not have to show it is actually true. This modest goal is the secret of the defence's success.
Key term: A defence shows only that "God exists" and "evil exists" are logically consistent, by supplying a possibly-true proposition under which both hold. A theodicy (e.g. Augustine, Hick, Griffin) makes the stronger claim of stating God's real reasons for permitting evil.
Before the technical machinery, it is essential to see why free will is supposed to be worth its terrible cost — for the defence claims not merely that free agents can do evil, but that a world containing free agents is good enough to justify God in creating them despite that risk. Plantinga states the value-claim plainly: a world containing creatures who are significantly free — able to perform morally significant actions, choosing good or evil without being causally determined — is, all else being equal, more valuable than a world containing no free creatures at all, or one containing only automata programmed to do right. Genuine love, genuine moral goodness, genuine relationship: none of these is possible for a puppet. A being that "loves" only because it is compelled does not love at all; goodness that is merely the playing-out of prior causes is not moral goodness in the fullest sense. If this is right, then God has a good reason to create free creatures, and the price of that decision — the possibility, even the likelihood, that some will choose evil — is the unavoidable shadow of a great good. The defence then needs to show only that it is possible that God could not secure the great good (freedom) without permitting the evil; the value-claim explains why God would accept that trade.
The free-will idea is ancient: Augustine (Lesson 3) already located the origin of evil in the misuse of creaturely freedom, arguing that God is not the author of evil because rational creatures freely turned from him. Plantinga's achievement is to take this intuition and reconstruct it with the precision of modern modal logic and possible-worlds semantics, so that it answers Mackie's logical argument on Mackie's own ground. Where Augustine told a historical story, Plantinga makes a purely logical claim about possibility.
Recall Mackie's most influential objection (Lesson 1): if it is possible to choose the good freely on one occasion, it is possible to do so on every occasion, so an omnipotent God could have created free beings who always freely choose the good — and, since he did not, he is either not omnipotent or not wholly good. Plantinga's reply targets the hidden assumption that God can actualise any logically possible world.
Notice how economical this reply is. Plantinga does not need to claim that transworld depravity is true, or even likely; he needs only that it is logically possible. Mackie claimed a strict inconsistency, and a strict inconsistency is destroyed by a single coherent counter-possibility, however improbable. This is the asymmetry that gives the defence its power and that beginners most often miss: against a charge of contradiction, the theist's burden is astonishingly light — produce one consistent story on which God and evil coexist — whereas against a charge of improbability (the evidential problem) the same light burden would be useless, since a bare possibility does nothing to raise a probability. Keeping this asymmetry in view is the key to understanding both why the defence triumphs in Lesson 1's arena and why it is powerless in Lesson 2's.
Plantinga frames this with possible-worlds semantics. A "possible world" is a complete, consistent way reality might have been; the actual world is one such world among infinitely many. Within this framework:
The key result is that God's omnipotence does not entail the power to actualise just any possible world. Worlds containing free creatures include the free creatures' own choices, which God does not control. God can create the conditions for free action; God cannot fix its outcomes while leaving it free. There may thus be logically possible worlds — including ones where free creatures always do right — that even an omnipotent God cannot bring about.
Plantinga sharpens the defence with the concept of transworld depravity. A person (more precisely, an "essence") suffers from transworld depravity if, in every feasible world in which that person exists and is significantly free, they would go wrong with respect to at least one morally significant action — that is, in any such world they would freely do at least one wrong thing.
Plantinga then makes the crucial possibility-claim: it is logically possible that every creaturely essence God could instantiate suffers from transworld depravity. If that possibility obtained, then no matter which free creatures God chose to create, and no matter what circumstances he placed them in, they would freely produce at least some evil. God could not then have created a world with significantly free creatures and no moral evil at all.
The argument does not assert that transworld depravity is actual — only that it is possible. And possibility is all a defence requires: if it is even possible that all creaturely essences suffer transworld depravity, then it is possible that God could not have made free creatures without evil, and the logical problem's claim of strict inconsistency fails. This is the precise hinge on which Plantinga answers Mackie.
The idea is easier to grasp through Plantinga's own style of illustration. Take a free agent — call him Curley — whom God is considering creating. For God to actualise a world in which Curley exists and freely does only right, it would have to be true that, in the circumstances God places him in, Curley would freely choose rightly on every occasion. But whether that counterfactual is true is not up to God; it is a fact about how Curley's freedom would in fact operate, a so-called "counterfactual of creaturely freedom". Suppose that, as a matter of fact, in every circumstance in which God could place him, Curley would freely go wrong at least once. Then there is no feasible world available to God in which Curley exists and is free and sins not at all — even though such a world is describable and so logically possible "in the broad sense". Now suppose this is true not just of Curley but of every possible free creature. Then, although a world of sinless free creatures is logically possible as a description, it is not within God's power to actualise, because actualising it would require the counterfactuals of freedom to be other than they are — and those are not God's to set. That supposition is transworld depravity, universalised, and its mere possibility is enough.
It is worth pausing on why Plantinga's argument was felt, across the discipline, to be decisive in a way few philosophical arguments are. Before Plantinga, it was widely assumed — by Mackie and many others — that the theist was caught in a flat contradiction and that the free-will reply was a feeble evasion. Plantinga's achievement was to take the reply, which Augustine had advanced as part of a historical narrative, and reconstruct it as a precise claim in modal logic that meets the atheist on the atheist's own deductive ground. He distinguished sharply between a world's being possible and its being actualisable by God, and showed that Mackie's argument illicitly assumed that omnipotence collapses the two. Once that assumption is exposed, the alleged contradiction simply dissolves, because the theist can consistently affirm all of: God is omnipotent; God is wholly good; evil exists; and God could not have actualised a world of free creatures with less evil. The proof that this set is consistent is what the logical problem demanded the theist provide, and what Plantinga provided. That is why Mackie conceded and why the discipline moved on.
Mackie had pressed the objection directly: "if God is omnipotent, he could have created beings who always freely choose the good; there is nothing logically impossible in a world of freely-good agents, so an omnipotent God could have actualised it." Plantinga's reply turns on a distinction Mackie blurs:
So Mackie's inference slides illegitimately from "logically possible that they always do right" to "God can make them always do right". A world can be possible in itself and yet not actualisable by God, because actualising it would require God to determine free choices, which is contradictory. Mackie himself, in The Miracle of Theism (1982), accepted that Plantinga's argument is successful against the logical problem, even while continuing to regard theism as improbable.
Everything in the free will defence depends on a particular, contested account of what freedom is. Plantinga assumes libertarian (incompatibilist) free will: an action is free only if it is not causally determined by prior states of the world, so that, in the very same circumstances, the agent could genuinely have done otherwise. On this view, God simply cannot both leave an action free and guarantee its outcome, because to determine the outcome would be to make the action unfree — hence Mackie's demand that God create free agents who always choose rightly asks for something contradictory.
The rival view is compatibilism, which holds that freedom is consistent with determinism. On the compatibilist account, an action is free if it flows from the agent's own desires and character without external compulsion — even if those desires and that character were themselves determined by prior causes. A kleptomaniac who cannot help stealing is unfree; an ordinary person who steals because they want to is free, even if their wanting was causally determined. If compatibilism is true, then there is no contradiction in God's creating agents who are both free and determined always to choose the good — God could give them a nature from which only good choices flow, and their good choices would still count as free because uncompelled. In that case Mackie's objection stands, and the free will defence collapses.
This is the deepest issue in the whole debate, and it is no accident that Mackie was himself a compatibilist. The free will defence does not refute compatibilism; it simply assumes its denial. Its success against the logical problem is therefore conditional: it shows that if libertarian freedom is coherent and valuable, then God and evil are not logically incompatible. A committed compatibilist will regard the defence as begging the question. Plantinga's reply is that libertarian freedom is at least coherent and arguably more valuable than its compatibilist counterpart — and since the logical problem claimed a strict contradiction, the theist needs only the coherence of the libertarian picture, not its proof, to block that claim. But the dispute over the nature of freedom remains genuinely unresolved, and a strong evaluation will identify it as the point on which the defence is most vulnerable.
Key term: Libertarian (incompatibilist) free will holds that a free act must be undetermined by prior causes — the agent could have done otherwise in the very same circumstances. Compatibilism holds that an act is free if it flows from the agent's own uncompelled desires, even if those desires were themselves determined.
The obvious gap is natural evil — earthquakes, disease, the fawn in the fire — which is not produced by human free choices. Plantinga addresses this with his extended free will defence: it is logically possible that natural evil is the result of the free actions of non-human persons, such as fallen angels, whose rebellion disorders nature. He does not claim this is plausible or actual; he claims only that it is not self-contradictory. Since a defence needs only logical possibility, even this speculative hypothesis discharges the logical problem with respect to natural evil. Most philosophers find the fallen-angel hypothesis ad hoc and implausible — but, importantly, implausibility is not inconsistency, so the objection has less force against a defence than it would against a theodicy.
Subscribe to continue reading
Get full access to this lesson and all 10 lessons in this course.