Why Does God Allow Suffering? The Real Argument

Why does God allow suffering? The direct answer is that He permits finite, temporal suffering within a world of free creatures and stable laws, for reasons sufficient to His nature and not always visible from ours, and that no amount of it ever becomes the settled rejection of good that would put His goodness in question. That answer is worth little to a reader who has not first seen how strong the objection is. The objection at its strongest is William Rowe's evidential argument, which claims probability rather than contradiction, and four serious replies stand against it: Alvin Plantinga's, John Hick's, Stephen Wykstra's and Gregory Boyd's. The Many Beings Framework names a fifth.

Key takeaways

• Why does God allow suffering: because what He permits is finite, and finite evil, however horrific, differs in kind from the absolute rejection of good His nature excludes.
• The live argument is evidential, not logical. William Rowe's 1979 case claims suffering makes God improbable, and an inference can be blocked without being contradicted.
• Four replies, four unadvertised prices. Plantinga's free will defence establishes possibility only. Hick's soul-making theodicy pays in universalism. Wykstra and Bergmann's skeptical theism pays in spreading scepticism. Boyd's open theism pays the settled future.
• Frame-Indexed Classical Theism blocks the inference rather than answering it. Where Plantinga supplies a possible reason and Hick an actual one, it argues that "we see no reason, so there is none" crosses two frames.

The weight of the question

Most people who search why does God allow suffering are not grading a syllogism. They are asking after a diagnosis, a phone call at four in the morning, a child who did not recover. It would be dishonest to treat such a person as though they were only an examiner, and worse to pretend that a page of philosophy is what they came for.

It is also true that the answer they are owed is a real one. An account of suffering that cannot survive examination will not survive grief either, because grief examines harder than any seminar does. The comfort that collapses under the first hard question was never comfort; it was a delay.

So the philosophy below is done properly and the rival positions are named fairly, which is the respect this question is owed. What follows is an argument rather than a consolation. The argument is the consolation, or it is nothing.

What does the problem of evil actually claim?

The problem of evil is the charge that the suffering we observe counts decisively against the existence of a God who is at once all-powerful, all-knowing and perfectly good. People ask it in two registers, one about suffering and one framed as why does God allow evil, and the literature treats these as one problem carrying two names. It comes in two forms, though, and the popular answers to why God allows suffering usually address the wrong one.

The logical problem of evil claims a strict contradiction. If God could prevent evil and would want to, then evil should not exist; it does exist; therefore no such God does. The evidential problem of evil claims far less and asks far more. It grants that the contradiction may be escapable, and argues instead that the sheer quantity and pointlessness of observed suffering make theism improbable.

Two further terms decide what counts as an answer. A defence shows only that God could have a morally sufficient reason for permitting some evil, without saying what the reason is. A theodicy goes further and says what the reason actually is. A defence that succeeds does not thereby explain anything, and a theodicy that fails can fail while the defence still stands.

That distinction is why the logical version faded from the literature and the evidential version did not. One was answered by a defence. The other asks for something a defence cannot supply.

Volume II works the two forms through side by side, with the older formulations set out in full.

Why is the evidential argument the hard one?

Because it never overreaches. William Rowe's 1979 statement of the evidential problem of evil has three steps, and each one is easier to grant than to refuse:

"There exist instances of intense suffering which an omnipotent, omniscient being could have prevented without thereby losing some greater good or permitting some evil equally bad or worse."

"An omniscient, wholly good being would prevent the occurrence of any intense suffering it could, unless it could not do so without thereby losing some greater good or permitting some evil equally bad or worse."

"There does not exist an omnipotent, omniscient, wholly good being."

The second premise is the one most believers hold already. It is close to a description of what goodness means. The whole weight therefore falls on the first, and Rowe supports it with a single case, chosen with great care:

"In some distant forest lightning strikes a dead tree, resulting in a forest fire. In the fire a fawn is trapped, horribly burned, and lies in terrible agony for several days before death relieves its suffering."

Notice what the example has been stripped of. There is no human choice in it, so no appeal to free will touches it. There is no soul being formed by it; a fawn is not building character as it dies. There is no witness to be improved by having seen it. Rowe is not reaching for the most emotive case available; he calls it "a familiar sort of tragedy, played not infrequently on the stage of nature."

Every reply below is an attempt to deal with one step, and the step states in a single sentence: from the fact that we can see no justifying reason for an instance of suffering, Rowe infers that probably there is none. A reader who cannot feel the pull of that step has not yet met the argument, and will answer something else instead.

The same argument is set out at length, with its critics, in Volume II.

Does free will explain why God allows suffering?

Partly, and Alvin Plantinga's version of the answer is generally regarded as the field's clearest success. His free will defence begins from a claim about value:

"A world containing creatures who are sometimes significantly free (and freely perform more good than evil actions) is more valuable, all else being equal, than a world containing no free creatures at all. Now God can create free creatures, but he cannot cause or determine them to do only what is right."

The second sentence carries the argument. A creature who can be determined to choose rightly is not choosing. Plantinga adds the formal move that made the defence decisive, which he calls transworld depravity: it is possible that every creaturely essence God could have created is such that, if it were created and left genuinely free, it would go wrong at least once. If that is so much as possible, then a world containing free creatures and no evil may be a world even omnipotence cannot make.

"The essential point of the Free Will Defense is that the creation of a world containing moral good is a cooperative venture; it requires the uncoerced concurrence of significantly free creatures."

The result is generally taken to have settled the logical problem of evil. What it costs is stated plainly by Plantinga himself. He offered a defence and not a theodicy, so it establishes that God could have a reason, and says nothing about what the reason is. It therefore does not touch the fawn. No creaturely will is in view in a lightning strike, and the defence has to be supplemented, usually by attributing natural evil to fallen angelic agency, which buys the answer at the cost of a further claim most readers will not grant.

Volume I argues freedom from the ground up and is free to download.

Is suffering how souls are made?

Yes, on John Hick's account, which takes the question the free will defence declines. His soul-making theodicy, drawn from Irenaeus rather than Augustine, denies that humanity was created finished and then fell from a height. On Hick's reading we were created immature, and the world is the environment in which we are brought to maturity. The label comes from the poet John Keats, who called the world a vale of soul-making.

The argument has real force. Courage cannot be formed in a world with nothing to fear. Compassion cannot be formed where no one suffers. Patience, generosity and forgiveness all require the very conditions a benevolent designer is accused of failing to remove. Hick adds that God maintains an epistemic distance from us, remaining less than obvious, so that turning toward Him is a free response rather than a compelled one. A world with God plainly visible in it would produce obedience, and obedience is not love.

This is a theodicy in the strict sense: it names the reason. That is its strength over a defence, and its exposure.

The price shows in the ratio. Some suffering plainly unmakes souls rather than making them, and the objection is not that Hick has no answer but that his answer is expensive. He holds that soul-making continues beyond death and ends, for everyone, in God. Universalism keeps the theodicy intact by removing the possibility of final loss, which means the doctrine of judgement has been adjusted to protect the account of suffering.

It also leaves Rowe's case roughly where it found it. A fawn dying alone in a forest is forming no character, and nobody is watching who might form some by seeing it. Hick answers the suffering that shapes persons, which is a great deal of suffering, and he is quietest exactly where the evidential argument does its work.

The series takes up final judgement directly in Volume II.

Can we tell that suffering is pointless?

Stephen Wykstra's reply says no: it grants Rowe the fawn and attacks the step that follows. His principle, CORNEA, sets a condition on what failing to see something entitles us to claim:

"On the basis of cognized situation s, human H is entitled to claim 'It appears that p' only if it is reasonable for H to believe that, given her cognitive faculties and the use she has made of them, if p were not the case, s would likely be different than it is in some way discernible by her."

In plainer words: not seeing a thing counts as evidence of its absence only when, had it been present, we would probably have seen it. Wykstra calls the bad version a noseeum inference, and it fails wherever the thing in question is the sort of thing we would miss.

Michael Bergmann developed the position into three theses, given here as the Stanford Encyclopedia's entry on skeptical theism states them:

"We have no good reason for thinking that the possible goods we know of are representative of the possible goods there are."

"We have no good reason for thinking that the possible evils we know of are representative of the possible evils there are."

"We have no good reason for thinking that the entailment relations we know of between possible goods and the permission of possible evils are representative of the entailment relations there are between possible goods and the permission of possible evils."

These are claims about the representativeness of our sample rather than about God, which is what makes the position economical. It requires no theodicy at all: it need not know why the fawn burned, only deny that we are positioned to rule out a reason.

The price is that scepticism does not stay where it is put. If our grasp of goods and evils is too unrepresentative to support Rowe's inference, critics ask why it stays good enough to support ordinary moral judgement, or to trust a divine promise. Skeptical theism buys silence, and silence is all it buys.

The framework's account of what our frame can certify is set out at the framework hub.

Would an open future clear God?

Gregory Boyd's answer is yes, and it moves the problem by changing what God knew. On open theism, the future free choices of creatures are not there to be known, so God did not foresee this history and select it from among alternatives.

Boyd builds his trinitarian warfare theodicy, in Satan and the Problem of Evil (2001), on six theses. Love must be freely chosen. That freedom is a genuine risk for God. Creatures are held responsible for how they use it. A creature's capacity for love is proportionate to its capacity for evil, so the power to bless and the power to devastate rise together. Freedom must be irrevocable, since a freedom God could withdraw at the moment of misuse would carry no responsibility. And that freedom is finite rather than endless, and redeemable while it lasts.

The gain is considerable. On this account the fawn and the child who did not recover are not items in a plan. Gratuitous evil is genuinely gratuitous, and the believer is not required to hold that each horror was selected for its role.

The price is the settled future, and with it the promises that rest on knowing it. Critics add a sharper point. Boyd's God retains the power to intervene and declines to, out of respect for an irrevocable freedom, which returns the argument to precisely the question it was built to escape: why this instance, why not that one.

Volume III treats divine knowledge and the shape of time in full.

Why do bad things happen to good people?

Because the question assumes a ledger, and Jesus refuses that ledger twice when it is put to Him directly. The phrase "why do bad things happen to good people" prices suffering against desert, and Scripture declines the exchange rate before anything else.

"Now there were some present at that time who told Jesus about the Galileans whose blood Pilate had mixed with their sacrifices. Jesus answered, 'Do you think that these Galileans were worse sinners than all the other Galileans because they suffered this way? I tell you, no! But unless you repent, you too will all perish. Or those eighteen who died when the tower in Siloam fell on them—do you think they were more guilty than all the others living in Jerusalem? I tell you, no! But unless you repent, you too will all perish.'" (Luke 13:1-5, NIV)

One case is political murder, the other a collapsed tower. In both, the inference from affliction back to guilt is refused outright. The same refusal appears when the disciples try it on a man who never had a chance to earn it.

"As he went along, he saw a man blind from birth. His disciples asked him, 'Rabbi, who sinned, this man or his parents, that he was born blind?' 'Neither this man nor his parents sinned,' said Jesus, 'but this happened so that the works of God might be displayed in him.'" (John 9:1-3, NIV)

A second assumption sits underneath the first: that "good" carries one settled sense across God and creatures. Jesus marks the difference himself: "No one is good—except God alone" (Mark 10:18, NIV). The right conclusion is not that nobody is wronged, which is the move most popular answers make and which cheapens question and questioner alike. Creaturely goodness is real, and injustice done to it is real injustice. What the verse denies is that the word carries the same content in both places.

Goodness and evil are handled as frame-indexed terms throughout Volume II.

How Frame-Indexed Classical Theism answers

Frame-Indexed Classical Theism (FICT) is a research program in analytic theology that resolves the classical dilemmas by diagnosing them as artefacts of a single formal error. Beings whose natures differ in kind apprehend reality through different frames, and predicates carrying value, time, agency or possibility are said across those frames analogically rather than univocally. The Many Beings Fallacy is predicating such a term univocally across frames and then treating the resulting clash as a discovery about God.

Three claims follow, and together they are this framework's answer to why does God allow suffering.

First, evil in our frame is real, culpable and grievous, and it is privative. It is the absence of a good that ought to be present rather than a substance in its own right. Aquinas puts it exactly: "Therefore it must be that by the name of evil is signified the absence of good", because "since being, as such, is good, the absence of one implies the absence of the other" (Summa Theologiae I, q. 48, a. 1, at New Advent). The tradition behind that reading is set out in our article on Thomism.

Second, finite evil never becomes absolute evil. Absolute evil would be a settled, final and everlasting repudiation of the good by a will capable of maintaining perfect righteousness, and no accumulation of creaturely horror inside history reaches it. The difference is one of kind and not degree, so scaling the quantity of suffering, the move on which the evidential argument depends for its force, does not change the category of the thing being scaled.

Third, finite evil is finalised at death. The finite is not trivial because it ends; it is decisive because it ends. The choices that constitute it resolve there into a fixed orientation, which is why this account does not shade into the claim that the pain was small.

Set against the rivals, the difference is what each one has to supply. Plantinga needed a possible reason and Hick needed an actual one; Wykstra needed a principle about seeing and Boyd needed an open future. Frame-Indexed Classical Theism needs none of those, because it denies Rowe his bridge rather than crossing it. The step from "we see no justifying reason" to "there is probably none" carries an expectation formed in the human frame into a verdict about the divine one, and nothing licenses the transfer.

What the framework gives up is univocity, which is a real cost, and it fails outright if a classical dilemma can be shown to turn on no frame-sensitive term at all.

Volume II, Resolving the Problem of Evil, carries this argument from its definitions to its conclusions.

How the answers compare

Each answers Rowe, and each pays somewhere.

Frequently asked questions

Is suffering a punishment from God?

Not as a general rule, and the inference from a particular affliction to a particular sin is one Jesus refuses outright in Luke 13:1-5 and John 9:1-3. Scripture does record judgement, and it distinguishes discipline, which is corrective and addressed to those God receives as children, from retribution. What it does not license is reading backwards from someone's suffering to their guilt.

What is the difference between a theodicy and a defence?

A defence shows that God could have a morally sufficient reason for permitting evil, without saying what it is. A theodicy says what the reason is. Plantinga offered a defence, which is why it settled the logical problem and left the evidential one standing; Hick offered a theodicy, which is why it explains more and costs more. Frame-Indexed Classical Theism offers neither, arguing instead that the inference from evil to improbability never acquires its licence.

Does the free will defence explain natural disasters?

Not on its own, and Plantinga did not claim it did. A lightning strike involves no creaturely choice, so the defence must be supplemented, historically by attributing natural evil to fallen angelic agency. The stronger line for natural evil is the stability argument: responsible agency requires a law-like world in which actions have predictable consequences, and a world of constant intervention would dissolve both the consequences and the responsibility.

What is skeptical theism?

Skeptical theism is the position that our grasp of possible goods, possible evils and the connections between them is too unrepresentative to support the claim that any particular evil is pointless. Wykstra's CORNEA and Bergmann's three theses are its standard statements, set out in the Stanford Encyclopedia's entry on the view. The standing objection is that a scepticism strong enough to block Rowe's inference looks strong enough to block ordinary moral reasoning as well.

Is frame-indexing just a way of saying God's ways are mysterious?

No, and the difference is testable. An appeal to mystery declines to say anything; Frame-Indexed Classical Theism makes a specific claim about the meaning of value terms across natures, delivers a specific verdict on a specific inference, and is falsified by any classical dilemma shown to involve no frame-sensitive term. Where Wykstra's skeptical theism restricts what we may conclude and Hick's soul-making theodicy supplies a reason, the framework identifies an equivocation and says exactly where it occurs.

Where to go from here

The reason the popular answers fail this reader is a mismatch. They answer the logical problem, carefully and often beautifully, while the reader in front of them is holding the evidential one. Anyone still asking why does God allow suffering after reading a page about free will has not misunderstood the page; they have noticed that it addressed a different argument.

Volume I, Resolving The Free Will Dilemma, lays the foundation and is free to download. Volume II, Resolving the Problem of Evil, is where this argument is made in full, from the definitions of good and evil through the evidential case to natural evil and human suffering. The series argues one position straight through rather than collecting essays that talk past each other, which is what a question this old has usually been denied.