The Problem of Evil

Apologetics usually meets the problem of evil as an attack from outside, posed by a philosopher or a sceptical student. Scripture meets it first from inside. Habakkuk is a prophet of Judah in the last years of the seventh century BC, watching the law paralysed and justice perverted in his own city (1:4), and his book opens not with an oracle to the people but with a complaint to God: 'How long, O LORD, must I call for help but You do not hear, or cry out to You, "Violence!" but You do not save?' (Habakkuk 1:2). When God answers that he is raising up the Babylonians to punish Judah, Habakkuk's second complaint is harder than his first. The anchor verse is the centre of it. God's eyes are too pure to look on evil, and yet he looks on while the wicked swallow those more righteous than themselves.

Notice the shape of that argument, because it is exactly the shape the philosophers would later give it. Habakkuk holds three things: God is holy, God is in charge of history, and the innocent are being devoured. He cannot see how all three can be true together. He does not resolve the difficulty by giving one of them up. He goes to his watchtower to wait for an answer (2:1), and the answer he receives is not an explanation but a promise and a command: the vision will come at its appointed time, and the righteous will live by faith (2:3-4). The book ends with a hymn sung in advance of any visible change. With the fig tree bare, the olive crop failed and the stalls empty (3:17), the prophet says: 'yet I will exult in the LORD; I will rejoice in the God of my salvation!' (Habakkuk 3:18).

Habakkuk is not alone. Job curses the day he was born and spends thirty chapters refusing his friends' explanations. Psalm 73 nearly stumbles over the prosperity of the wicked. Psalm 44 tells God to wake up. Jeremiah asks why the way of the wicked prospers (12:1). And at the centre of the New Testament the Son of God cries the opening line of Psalm 22 from the cross. A student should take two things from this before any philosopher is named. First, the Bible does not regard the question as impious; God rebukes Job's friends, not Job, for what they said about him (Job 42:7). Second, the Bible never gives the full answer to it. It gives a great deal that bears on the question, and it ends by asking for trust.

The philosophical form is old as well. The Christian writer Lactantius, around the year 313, attributes to Epicurus a fourfold dilemma: God either wishes to take away evils and cannot, or can and does not wish to, or neither wishes nor is able, or both wishes and is able, and if the last, where do evils come from? (On the Anger of God, chapter 13). David Hume put it into the mouth of his sceptic Philo in Dialogues Concerning Natural Religion, published in 1779 after his death, and his wording is the one most often quoted: 'Is he willing to prevent evil, but not able? then is he impotent. Is he able, but not willing? then is he malevolent. Is he both able and willing? whence then is evil?' Those three questions are the spine of everything in sections 2 and 3.

One more preliminary, and it is the most important. The objection that costs people their faith is rarely the philosopher's version. It is the death of a child, a war that took a village, a disease that came to the person who prayed hardest, a pastor who abused his trust and was never stopped. Bart Ehrman, a New Testament scholar whose work appears elsewhere in this track, has written that what finally ended his Christian faith was not textual criticism but suffering; his book on the subject is God's Problem (2008). The argument and the wound usually arrive together. This lesson deals mainly with the argument, because the argument deserves an answer and a minister must be able to give one. It deals with the wound only enough to tell you where the argument stops being the right tool.

In 1955 the Australian philosopher J. L. Mackie published a short article in the journal Mind called 'Evil and Omnipotence'. It set the terms of the debate for twenty years, and it deserves to be stated as strongly as he stated it. Mackie argued that the theist holds three propositions: God is omnipotent, God is wholly good, and evil exists. Add two premises that look like no more than definitions: that a good being always eliminates evil as far as it can, and that there are no limits to what an omnipotent being can do. Then the five cannot all be true. The problem, Mackie said, is not that theism lacks evidence. It is that its central beliefs contradict one another, so that holding them is not merely unsupported but irrational. If he is right, no amount of evidence for God could help, because a contradiction cannot be true however much evidence there seems to be for it.

Mackie anticipated the obvious reply. Christians say that God permits evil for the sake of free will, because a world of free creatures who sometimes do wrong is better than a world of puppets. Mackie's answer was sharp. If freedom is compatible with sometimes choosing the good, it is surely compatible with always choosing the good. Many people freely choose the good on many occasions. Why could God not have made creatures who freely chose the good on every occasion? An omnipotent God, on this view, had a better option available and declined it, and that is hard to square with perfect goodness.

The reply that turned the debate came from Alvin Plantinga, in The Nature of Necessity and in a shorter and more readable book, God, Freedom, and Evil, both published in 1974. Plantinga began with a distinction that a student must keep clear. A theodicy claims to give God's actual reasons for permitting evil. A defence claims much less: only that there is a possible reason, one that might be true for all anyone can show, under which God and evil are consistent. To defeat a charge of contradiction you do not need to know the true story. You need only to describe one possible story in which all the propositions are true together. If such a story is possible, there is no contradiction.

His story turns on freedom in the strong sense, usually called libertarian freedom: an action is free only if nothing, including God, determines that the creature does it. If that kind of freedom has great value, then even an omnipotent God cannot make it the case that a free creature freely does right, because if God makes it the case, the act is not free. Plantinga called the assumption that God could create any possible world 'Leibniz's lapse'. Some possible worlds, such as one in which every free creature always freely does right, may not be worlds God can bring about, because whether they obtain depends partly on what free creatures would do. It is possible, he argued, that every creature God could make would go wrong at least once if made free, a condition he named transworld depravity. If that is even possible, then it is possible that God could not create a world containing moral good without also permitting moral evil. Mackie's contradiction disappears.

Mackie's objection about earthquakes and diseases, which no one chooses, was answered in the same spirit. Plantinga noted that it is possible that natural evils are the work of free and fallen non-human persons, as Christian tradition has long held of Satan and his angels, and since only possibility is needed, that is enough to remove the contradiction. Many readers find that suggestion unconvincing as an account of what actually happens. That is to mistake what it is for. As a theodicy it would need defending; as a defence it only needs to be possible.

The result is as close to a verdict as philosophy of religion has reached. Most philosophers who work in the field, theists and atheists alike, now judge that the logical problem in Mackie's form has been answered. William Rowe, whose own argument is the subject of the next section, wrote in 1979 that he did not think anyone had succeeded in showing that God and evil are logically inconsistent, and Mackie himself, in The Miracle of Theism (1982), allowed that the problem does not show the central doctrines of theism to be strictly inconsistent with one another. An honest apologist should say this plainly, and should also say what it does not mean. It does not mean the problem of evil has been solved. It means one version of it, the claim of outright contradiction, no longer carries the weight Mackie put on it.

There is a live disagreement inside the church at this point, and it should not be hidden. Plantinga's defence depends on libertarian freedom. Many Reformed theologians hold that human freedom is compatible with God's ordaining of everything that comes to pass, as Westminster Confession 3.1 says, while denying that God is the author of sin. They cannot use Plantinga's story as it stands, and they do not need to. They argue instead that God permits and ordains evil for a greater good that he knows and we do not, pointing to Joseph's words to his brothers, 'what you intended against me for evil, God intended for good' (Genesis 50:20), and to the cross, where the worst act in history was done by lawless hands according to God's set plan and foreknowledge (Acts 2:23). John Feinberg's The Many Faces of Evil (1994) argues that each theology must answer the problem on its own terms rather than borrow another's. Arminian and Reformed Christians agree that God is good and does no evil. They differ over how freedom and providence fit together, and so they answer Mackie in different ways.

If the logical problem has been answered, why does the problem of evil remain the most serious objection to Christian belief? Because a stronger argument replaced it. In 1979 William Rowe published 'The Problem of Evil and Some Varieties of Atheism' in the American Philosophical Quarterly. He set out not to prove a contradiction but to show that evil makes the existence of God improbable. He chose an example deliberately free of human sin. Lightning strikes a dead tree in a distant forest; a fire breaks out; a fawn is trapped, horribly burned, and lies in agony for several days before it dies. No human being knows. No one learns from it, is tested by it, or grows through it.

His argument has three steps. First, there probably exist instances of intense suffering which an all-powerful, all-knowing being could have prevented without thereby losing some greater good or permitting some evil equally bad or worse. Second, a wholly good and all-knowing being would prevent any intense suffering it could, unless it could not do so without losing some greater good or permitting some worse evil. Third, therefore, there probably does not exist an all-powerful, all-knowing, wholly good being. The second premise is one almost every theist accepts; it simply says God does not permit suffering for nothing. Everything turns on the first. Rowe's support for it is plain: we have looked for the good that would justify the fawn's suffering, and we cannot find one, and it seems reasonable to conclude from this that there is none.

Rowe's argument is stronger than it first looks for three reasons. It does not need to show that every evil is pointless; one pointless evil is enough, and the world contains a great many candidates. It is not refuted by showing that some suffering produces good, since no one denies that. And it is stated by someone who was careful to be fair. Rowe called himself a friendly atheist: he held that some theists might be entirely rational in their belief, since they might have other grounds for it, while maintaining that the evidence from suffering points the other way. In later writing he added a second case, a real one, the murder of a five-year-old girl in Michigan, and asked what good could possibly be served by permitting it. A student who has sat with a bereaved family knows that this is the form in which the question is actually lived.

Two other versions should be known. Paul Draper, in a 1989 article in the journal Nous titled 'Pain and Pleasure: An Evidential Problem for Theists', compared two hypotheses: that the world is governed by a good God, and what he called the hypothesis of indifference, that nothing in nature is the result of benevolent or malicious action by supernatural persons. He argued that the actual pattern of pain and pleasure across living things, the way pain often serves survival and often does not, is much more what we should expect on the second hypothesis than on the first. And Marilyn McCord Adams, an Episcopal priest and a philosopher, pressed the problem from inside the faith in Horrendous Evils and the Goodness of God (1999). She argued that the general goods appealed to by most theodicies, such as freedom or a law-governed world, do not show that God is good to the particular person whose life has been swallowed by a horror. A God who balanced the books for the world as a whole while one child's life was destroyed would not be good to that child.

Adams belongs in this section rather than the next because her challenge does not go away when Rowe's is answered. She held that God is good, and that the only adequate answer is that God will defeat each horror within the life of the one who suffered it, by an intimacy with God so great that the sufferer would not wish the horror out of his story. She also drew the conclusion that God must finally save every person, which most Christian traditions have not affirmed, and which lesson 2 of this module will treat on its own terms. Whatever one makes of that conclusion, her test is worth adopting: an answer to the problem of evil must be an answer that could be said, in time, to the person it is about.

The first reply is sceptical theism, and it goes after Rowe's first premise directly. In 1984 Stephen Wykstra argued, in the International Journal for Philosophy of Religion, that the step from 'we cannot see a reason' to 'there probably is no reason' is valid only if we would expect to see the reason if there were one. He called this the condition of reasonable epistemic access, CORNEA for short. Looking in my garage and seeing no car is good evidence there is no car; looking out of my window and seeing no bacteria on the wall is no evidence at all, because I would not see them if they were there. The question is which kind of case the fawn is. If God exists and his knowledge exceeds ours by more than a parent's exceeds an infant's, Wykstra argued, we should positively expect that many of his reasons will be beyond us. Michael Bergmann and others have developed the argument in detail. Its biblical home is the end of Job, where God answers the sufferer with questions about the foundations of the earth (Job 38:4) and never tells him about the conversation in heaven in chapter 1, and Isaiah's 'My thoughts are not your thoughts' (Isaiah 55:8).

Its critics have a serious answer. If we cannot judge that the fawn's suffering is pointless, can we judge that any suffering is pointless, and if not, why should we prevent any? A sceptical theist who sees a child drowning, on this objection, ought to wonder whether God has some reason for allowing it and stand still. Michael Almeida and Graham Oppy pressed a version of this in 2003, and others have argued that the same scepticism undermines confidence in revelation, since God might have reasons we cannot see for allowing us to be misled. Defenders reply that our duties are given to us by God's revealed commands and not by our estimate of the final outcome, so that knowing we must save the child does not require knowing what God will bring out of every case, and that scepticism about God's reasons for permitting evil is not scepticism about whether God keeps his promises. The exchange is unfinished, and a student should not present either side as having won.

The second family of replies tries to supply reasons rather than deny we can see them. The free will theodicy, in its fuller form, says that moral evil is the cost of a world in which love and goodness are freely given, and Plantinga's defence becomes a theodicy if one adds that this is in fact God's reason. Soul-making, associated with Irenaeus of Lyons in the second century and with John Hick's Evil and the God of Love (1966), says the world is designed not as a comfortable shelter but as a place where creatures grow toward maturity, so that courage, compassion and trust are possible only where real danger and loss are possible too; its texts include Romans 5:3-4 and Hebrews 2:10, where the author of salvation is made perfect through suffering. Richard Swinburne, in Providence and the Problem of Evil (1998), adds that a world governed by regular natural laws, in which fire burns whether or not anyone deserves it, is the condition of responsible action at all, because only in such a world can our choices have predictable effects. Each of these explains a great deal. The common objection to all of them is the fawn: much suffering seems too great, too random or too far from any human growth for these goods to cover it.

Christian theology adds something none of these general theories contains, and the apologist should not be shy about it. Plantinga himself, in a 2004 essay, argued that the best possible worlds might all include the incarnation and atonement, the greatest good there is, and that those require a fallen world to redeem; the Easter liturgy's ancient cry 'O happy fault' (in Latin, O felix culpa) points the same way. More broadly, the Christian claim is not only that God has reasons but that God has entered the suffering and will end it. The one who cried 'My God, My God, why have You forsaken Me?' (Mark 15:34) is the one Christians call God the Son, and the creation that 'has been groaning together in the pains of childbirth until the present time' (Romans 8:22) is promised a redemption in which 'He will wipe away every tear from their eyes' (Revelation 21:4). Paul's measure is explicit: 'I consider that our present sufferings are not comparable to the glory that will be revealed in us' (Romans 8:18). N. T. Wright's Evil and the Justice of God (2006) argues that the Bible's answer is less an explanation of evil than an account of what God has done and will do about it.

The third reply concerns the whole argument rather than one premise, and Rowe himself gave it its name. He observed that a theist can run his argument backwards. If I have good grounds for believing that God exists, from the arguments of module 2, from the resurrection in module 3, or from the experience of God, then I have good grounds for denying that the fawn's suffering is pointless, even though I cannot say what the point is. Rowe called this the G. E. Moore shift, after the philosopher who answered sceptics about the external world by saying he was more certain he had hands than he was of any premise in their arguments. It is a legitimate move, and its strength depends entirely on the strength of the other evidence. It does not show what the reason is. It shows that the problem of evil must be weighed against everything else, not alone.

One more reply is often made and should be handled with care. C. S. Lewis described how his own early argument against God was that the universe seemed so cruel and unjust, and how he came to ask where he had got the idea of just and unjust in the first place. The point, developed in the moral argument of module 2, is that the strongest form of the objection assumes that some things really are evil, and that a purely natural world struggles to say why. That point is real. It does not, however, answer Rowe, because Rowe's argument works from inside theism's own standards: if God is good, as the Christian says, then by the Christian's own measure there should be no pointless suffering. Use the moral argument against a sceptic who says that nothing is really evil. Do not use it to avoid the question of a sceptic who has granted that the fawn's agony is evil and is asking you why God allows it.

Put the results together honestly. The charge of contradiction has been met; a believer is not irrational in holding that God is good and that evil exists. The evidential argument has not been settled either way. Sceptical theism makes a strong case that we are in no position to say there is no reason for the fawn, and its critics make a strong case that it proves too much. The theodicies supply real goods that would justify much suffering, and none of them plausibly covers all of it. The Moore shift is as strong as the case for God from everything else, which is why this module comes after modules 2 and 3. The distinctively Christian claims, incarnation, cross, resurrection and new creation, do not explain why any particular evil happened, and they change the question from whether God cares to whether God can be trusted when he has not explained.

So say what you know and say what you do not. A Christian can say that God is not the doer of evil and is its enemy; that evil is real and not an illusion; that suffering is not proof of the sufferer's guilt, as Jesus says in Luke 13 and John 9; that God has entered human suffering in Christ; and that 'The last enemy to be destroyed is death' (1 Corinthians 15:26). A Christian cannot honestly say why this child died, or that any of the theories in section 4 removes all difficulty, or that a person who still finds the problem crushing has failed to understand it. Careful Christian philosophers have said openly that the problem of evil remains hard for them. Eleonore Stump's Wandering in Darkness (2010) works through Job and other biblical narratives because she judged that stories show what arguments cannot.

This is the place to name the difference between the lectern and the bedside. Everything in this lesson belongs at the lectern: in a class, a debate, a careful conversation with a sceptic, or a sermon on a passage like Habakkuk. Almost none of it belongs in the first days after a death. The Pastoral Ministry and Care track has a lesson on exactly that, Suffering Without Easy Answers, in its module Long Care. It begins with Job's friends, who did their best work in the seven days they sat in silence, and it sets out what not to say and how each of the answers above can comfort or wound. Read it alongside this one. The apologist who cannot tell when to stop arguing has learned the arguments and missed the point of them.

Finally, notice what the problem of evil looks like where you live, because it does not look the same everywhere. In much of the West it arrives as Rowe's fawn or as a news report of a distant disaster, and the question is whether God exists. In much of Africa and Asia it arrives as the question John 9 records, who sinned, and the answer offered is often a person: a witch, a jealous relative, an offended ancestor. That answer can lead to the accusation and abuse of the weak, often elderly women and children, and module 5 returns to it. Jesus' refusal to name a culprit in John 9 is an apologetic answer and a pastoral protection at the same time. In places of persecution the question is Habakkuk's own, how long, and the psalms of lament are more useful than any philosopher. Know which question you are being asked before you answer it.

The capstone of this track asks you to write a defence of the faith against the three objections most often raised where you live. This module is where most of that material will come from, so each lesson adds one sheet to a section of your notebook headed The hardest objections. Use the same six lines on every sheet: the objection at its strongest, in words its best defender would accept; the objection as people actually ask it where you live; what you can honestly answer; what your answer does not show; the text you stand on; and when you stop arguing and simply stay.

For this sheet, write the problem of evil twice under the first line. Once in Mackie's form, as a charge of contradiction, and once in Rowe's form, as an argument from probability, each in three numbered steps in your own words. Under the second line, go back to the page you wrote in the inquiry and copy the question in the words the person used. Then decide which of the two forms it is closest to, or whether it is the who-sinned question of John 9 instead, and write that down.

Under the third and fourth lines, write no more than half a page in plain language you could say aloud to a thoughtful friend. Say what the free will defence achieves against Mackie and what it does not achieve against Rowe. Name one reply to Rowe you find strongest and the best objection to it. Under the fifth line, write out Habakkuk 1:13 and 3:17-18 in full. Under the sixth, write one sentence that tells you when you will stop explaining and start sitting with someone, and the title of the Pastoral Care lesson you will reread before you visit. Then read your half page aloud to one other believer and ask which sentence sounded least honest. Rewrite that sentence.