There is a sentence spoken at gravesides, in hospital waiting rooms, and across kitchen tables where people sit with hands wrapped around mugs of cold tea and nowhere to put their grief. The sentence is this: everything happens for a reason. It is offered as consolation. It functions, in practice, as a theological claim so ambitious, so audacious in its implications, that most of the people who speak it have never paused to examine what would have to be true for the words to mean anything at all. This essay is that examination. The claim under test is specific and traditional: that God exists, that he is omnipotent, omnibenevolent, and omniscient, and that suffering therefore occurs within a framework of divine purpose that renders it, if not good, then at least explicable and tolerable. That claim will be weighed against the evidence. The conclusion will be stated plainly at the end, and it will not be comfortable, but it will be honest, which is the least the subject deserves.
The claim that everything happens for a reason is not an observation about physical causality. Scientists believe in causality too; every event has prior physical causes, and that chain of cause and effect is fascinating and well-studied. The sentence, as deployed in the hospital waiting room, means something categorically different: it means that a conscious agent with purposes and the power to achieve them arranged this event, that the event therefore has meaning beyond its brute occurrence, and that the suffering it occasions is in some sense justified by the plan it serves. Strip away the soft language, and what you are left with is the central claim of theodicy: God is real, omnipotent, omnibenevolent, and omniscient, and suffering exists within a framework of divine purpose. This is a claim that can be examined. When that examination is carried out honestly, the claim does not survive. What follows is the evidence for that verdict.
Steven Pinker identified the underlying architecture of the problem with characteristic precision: “A major breakthrough of the Scientific Revolution, perhaps its biggest breakthrough, was to refute the intuition that the universe is saturated with purpose. In this primitive but ubiquitous understanding, everything happens for a reason, so when bad things happen, accidents, disease, famine, poverty, some agent must have wanted them to happen.” That is exactly the structure at stake here. To say everything happens for a reason, in the theological sense, is to say that someone wanted childhood leukaemia to happen, that someone wanted the 2004 Indian Ocean tsunami to happen, that someone wanted the Holocaust to happen. If God is the author of the plan, then God is the author of those events. Theodicy is the long, elaborate, and ultimately unsuccessful attempt to make that authorship morally acceptable.
The Old Dilemma, Still Unanswered
Epicurus, writing in the third century BCE, formulated the problem of evil with a precision that has never been improved upon. The argument, as it has come down to us through Lactantius, runs as follows: either God is willing to prevent evil but not able, in which case he is not omnipotent; or he is able but not willing, in which case he is malevolent; or he is both able and willing, in which case there is no evil; or he is neither able nor willing, in which case he is not worth calling God at all. The dilemma is not a rhetorical flourish. It is a logical structure, and its force has not diminished over twenty-five centuries of theological commentary. You cannot have a God who is simultaneously all-powerful, all-knowing, and wholly good, and also a world that contains the quantity and quality of suffering that this one demonstrably does. The three attributes are mutually incompatible with the observable evidence. Every theodicy in the history of theology is an attempt to find an exit from the trap that Epicurus built two and a half millennia ago. None of them has succeeded. What they have found, instead, are increasingly sophisticated ways of quietly abandoning one of the three attributes while pretending to maintain all three.
John Stuart Mill, a man not given to casual moral outrage, arrived at the same conclusion by a different route. He recorded that his father James Mill’s rejection of religious belief “was not, as many might suppose, primarily a matter of logic and evidence: the grounds of it were moral, still more than intellectual. He found it impossible to believe that a world so full of evil was the work of an Author combining infinite power with perfect goodness and righteousness.” The significance of this observation extends well beyond the Mill household. The problem of evil is not merely an intellectual puzzle for professional philosophers; it is a moral verdict that thoughtful, feeling human beings have been reaching independently, across cultures and centuries, whenever they have looked at the world honestly and then looked at the theology that claims to account for it. The theology does not account for it. The various defences that theology has constructed deserve to be examined one by one, not because they are without intellectual interest but because the moral cost of each failed defence is real and falls on real people.
Before examining those defences, it is worth being precise about what they would need to accomplish. A successful theodicy would have to show that the existence of an omnipotent, omnibenevolent, and omniscient God is logically compatible with the specific suffering we observe in this world, not with suffering in the abstract, but with the particular, irreducible, and often grotesquely disproportionate suffering that fills the historical and medical record. Showing that some suffering might, in principle, serve some good purpose in some possible world would not suffice. The task is to account for this world, with this distribution of suffering, including its most extreme and apparently purposeless instances. That is the standard against which each theodicy must be measured, and it is the standard against which each one fails.
The Free Will Defence: A Partial Excuse for Total Disaster
The free will defence is the most venerable and the most frequently invoked of the theodicies. Its logic is straightforward: God permits evil because the alternative would be to create automatons rather than morally free agents, and a world containing genuinely free creatures who sometimes choose evil is better, in some ultimate accounting, than a world of puppets who cannot choose at all. Alvin Plantinga has developed this argument with genuine philosophical rigour, and it deserves to be taken seriously before being found wanting.
The first and most obvious difficulty is that the free will defence accounts only for moral evil, meaning the suffering caused by the choices of human beings, and not at all for natural evil, meaning the suffering caused by earthquakes, floods, cancers, parasites, and diseases. The 2004 Indian Ocean tsunami killed approximately 227,000 people. No human being chose to generate that tsunami. No act of free will produced it. The tectonic plates did not consult anyone’s moral agency before releasing their energy. The children drowned in that catastrophe were not drowned by a free moral agent choosing badly; they were drowned by a geological event that an omnipotent God could have prevented at any moment without compromising anyone’s freedom to choose anything whatsoever. The free will defence, however elegantly constructed, is simply irrelevant to this entire category of suffering. It cannot bear the weight that is placed upon it, because the weight includes a class of events to which the defence is entirely inapplicable.
Some theologians have attempted to extend the free will defence to natural evil by attributing it to the free choices of fallen angels or demonic powers, an argument that introduces a further layer of unsubstantiated metaphysical claim without resolving the underlying problem, since an omnipotent God could equally have restrained such powers without compromising their freedom in any morally significant sense, precisely as a human court restrains an offender without abolishing the concept of free will. The extension of the defence to cover natural evil therefore either generates the same dilemma at a new level or reduces to the claim that God is constrained by principles he did not himself set, which is not the God of classical theism.
The second difficulty is more fundamental and concerns the nature of omnipotence itself. The free will defence assumes that God cannot create beings who are both genuinely free and reliably inclined toward good, because such a combination is logically impossible. But this assumption is far from obvious. We can readily conceive of a world in which free agents exist and happen, through their own choices, to produce far less catastrophic suffering than the inhabitants of this world have produced. The range of possible worlds is vast, and within that range there are presumably many in which free agents exist and the Holocaust does not. An omnipotent God could have actualised any of those worlds instead of this one. The question is not whether freedom and goodness are logically compatible in principle, because they plainly are. The question is why a God who could create any possible world chose to create one in which the Holocaust, the Inquisition, the Thirty Years’ War, and the Rwanda genocide are among the historical facts. The free will defence does not answer that question; it systematically redirects attention away from it.
There is also a third difficulty, which is the most uncomfortable of all, because it reveals something about what the free will defence requires us to say about God’s own choices. God, in classical theology, is perfectly good and perfectly free. He is not compelled to do good by any external force; he chooses it freely. His perfect goodness is therefore compatible with his perfect freedom. If that is so, then perfect goodness and perfect freedom are logically compatible, which means that an omnipotent God could have created free human beings who were, like him, reliably good without being mere automatons. The free will defence requires us to believe simultaneously that God’s freedom and goodness are compatible, and that no free creature’s freedom and goodness could be guaranteed to be compatible. The position is not incoherent in every possible formulation, but it requires very careful philosophical argumentation to avoid the contradiction, and most of the people who deploy the free will defence at a graveside are not offering carefully formulated philosophical argument. They are offering a thought-terminating comfort, and that comfort does not, on close inspection, hold.
Soul-Making: The Child as Raw Material
The second major theodicy is associated with the theologian John Hick, though its roots reach back to the writings of Irenaeus in the second century. Hick’s soul-making theodicy begins with a different picture of humanity than the Augustinian tradition provides. Rather than imagining human beings as originally perfect creatures who fell from grace, the Irenaean tradition imagines humans as created immature, as morally and spiritually underdeveloped beings who require the encounter with challenge, difficulty, and suffering in order to grow into the full realisation of their potential. On this view, suffering is not a defect in the divine plan; it is a constitutive feature. This world is not a paradise from which we have fallen; it is a vale of soul-making, a kind of moral gymnasium in which the resistance of difficulty is precisely what enables spiritual growth.
Hick’s theodicy is philosophically interesting, and it has the merit of taking seriously the intuition that struggle and adversity sometimes produce genuine moral depth in those who survive them. This much is empirically supportable. People who have suffered greatly sometimes emerge with a compassion and a wisdom that comfortable lives do not easily generate. The empirical kernel is real. The theological structure built upon it is another matter entirely.
Consider what the soul-making theodicy requires us to say about a child who dies of cancer at the age of four. The child has not been given sufficient time to undergo the spiritual development that the soul-making process is supposed to produce. The child has experienced suffering without the developmental benefit the theodicy postulates. If the defence is to work at all in such cases, it must be supplemented by some claim about an afterlife in which the soul-making process continues or is somehow retroactively completed, a claim that introduces a new layer of assertion without any supporting evidence, and that begins to look less like a defence of God’s goodness and more like an infinite regression of justifications, each one requiring another to prop it up. The soul-making theodicy quietly imports eschatology every time it is applied to cases of early death, and it cannot stand without that import.
There is also the question of proportion. The soul-making defence might conceivably justify a world containing hardship, loss, and the kind of suffering that, in reasonable doses, does appear to produce moral growth. It cannot, without grotesque special pleading, justify the scale and the randomness of actual suffering in this world. The Black Death killed approximately a third of Europe’s population in the fourteenth century. If that catastrophe was a soul-making exercise, its efficiency was extraordinarily low: most of its subjects died too quickly, too young, and in too much agony to have derived any meaningful spiritual benefit from the experience. The Holocaust, similarly, did not function as a soul-making exercise for the six million murdered. It functioned as a mechanism of systematic annihilation. To say that it served a divine educational purpose is not theology with any claim to moral seriousness; it is an insult to the dead, and it is an insult that theology has not yet had the honesty to withdraw.
What the soul-making theodicy reveals, when pressed to its limits, is a picture of God as a pedagogue of extreme methods who could perfectly well achieve the same educational outcomes through gentler means, since he is, by hypothesis, omnipotent, but who chooses not to. This picture is not of a loving father. It is of a being whose commitment to a particular educational philosophy outweighs his apparent interest in the welfare of his students, many of whom do not survive the curriculum. The argument does not vindicate divine goodness; it presents a picture of divine indifference that is difficult to distinguish from the picture that atheism presents, except that atheism does not require us to call the indifference a plan.
The soul-making theodicy also has an uncomfortable implication that its proponents rarely spell out. If suffering is the necessary condition of moral development, then an omnipotent God who could eliminate suffering but chooses not to, on the grounds that suffering produces better souls, is a God who has decided that the quality of the product justifies the cruelty of the process. This is precisely the reasoning that we reject, as a matter of settled moral principle, whenever a human institution employs it. We do not permit schools to torture pupils on the grounds that the ordeal produces stronger character. We do not permit states to inflict gratuitous suffering on citizens on the grounds that adversity builds resilience. The fact that the institution deploying this reasoning is divine does not change the moral character of the reasoning; it simply places the reasoning beyond reach of the accountability mechanisms that we apply in every other context. That is not a defence of the reasoning. It is a description of its insulation from scrutiny.
The Greater Good: When God Becomes a Utilitarian
A third class of theodicy appeals to consequences: individual suffering, however terrible, serves some greater good that a limited human perspective cannot perceive. God sees the whole of the pattern; we see only individual threads. What appears, from our vantage point, to be a monstrous evil is, from the divine vantage point, a necessary element in a larger design of good. This is the theodicy that most directly underwrites the phrase “everything happens for a reason,” because it asserts precisely that: every event, including every catastrophe, is rationally ordered toward an outcome that justifies it.
This defence has a certain structural plausibility as long as it remains abstract. The moment it is applied to a specific case, its moral character becomes apparent. If the death of a four-year-old from leukaemia serves a greater good, then either that greater good was achievable without the child’s death, in which case an omnipotent God who chose this route had options and chose suffering when he need not have, or it was not achievable without the child’s death, in which case God’s omnipotence is being quietly withdrawn from the argument. These are the only two possibilities available, and neither is consistent with the traditional attributes of God that the theodicy is meant to defend.
The appeal to greater goods that we cannot perceive has a further structural problem: it is not a defence at all, in the philosophical sense. A genuine defence of a position involves providing a reason to believe that the position is true. The greater good theodicy, in its strong form, says only that we cannot rule out the possibility that there is a good reason for suffering, even though we cannot identify what that reason might be. Unfalsifiability is not a virtue in this context. A claim that is consistent with every possible observation, which the claim that suffering serves invisible greater goods certainly is, has no explanatory power whatsoever. If God’s goodness is compatible with any amount of suffering, provided only that we stipulate invisible compensating goods, then the attribute “omnibenevolent” has been evacuated of all content. It no longer constrains what we expect to observe. The theodicy saves the word while abandoning the substance, which is not a philosophical achievement but a philosophical sleight of hand.
There is also a deeply uncomfortable parallel with a certain kind of secular authoritarianism that theology rarely acknowledges. The logic “individual suffering is justified by the larger plan” is precisely the logic that has been used to justify every political atrocity committed in the name of historical necessity, from Stalinist collectivisation to the ethnic cleansing of populations deemed to stand in the way of national destiny. Theology is not responsible for political authoritarianism, but the structural similarity of the argument is worth making explicit, because it reveals that “greater good” reasoning, deployed without constraint, is capable of justifying literally anything. A theodicy that validates any possible state of affairs, no matter how monstrous, by appeal to invisible consequences has not solved the problem of evil; it has declined to take the problem seriously.
The greater good theodicy also encounters a specific problem when it is applied to the suffering of non-human animals. Before any human being existed, before any soul was available to be made, before any act of free will had been committed, the natural world was already a theatre of suffering on a colossal scale. The geological record is a record of mass extinctions, of organisms dying in conditions of extreme distress, without any theological framework, without any soul to be improved, without any human moral agent whose freedom needed to be preserved. The greater good that all of this pre-human suffering was meant to serve has never been specified by any theodicist, because there is no candidate good available that is not transparently post-hoc rationalisation. The suffering of the pre-Cambrian biosphere preceded any possible human recipient of any possible benefit, which makes the appeal to compensating goods structurally empty before the argument even reaches the question of human suffering.
You can read more about the moral architecture of this problem in the companion piece on this site, The Silent Observer: God and Atrocity, which examines the theological implications of divine inaction in the face of historical catastrophe.
The Inscrutable Plan: When Theology Becomes a Shrug
The fourth major theodicy is arguably the most honest of the four, which is not to say that it is adequate. This is the appeal to divine inscrutability: God’s ways are not our ways; his plan surpasses human understanding; and we should not presume, with our finite and fallen minds, to judge what an infinite and perfect mind has ordered. This is the theodicy of the Book of Job, in which God’s response to Job’s anguished demand for an explanation is, essentially, “Where were you when I laid the foundations of the earth?” It is the theodicy of the mystical tradition, which counsels acceptance of the incomprehensible divine will. It is also, in its popular form, the theodicy of “God has a plan” and “his ways are mysterious,” phrases that function as conversation-closers rather than arguments.
The inscrutability theodicy at least has the virtue of not pretending to explain what it cannot explain. It concedes, implicitly, that no satisfactory explanation of suffering is available to human reason, and it counsels trust rather than understanding. For many believers, this counsel carries genuine weight: the relationship with God is personal, experiential, and not reducible to propositional argument, and the demand for rational justification of divine action may seem to misunderstand the nature of faith. This is a position that deserves to be understood on its own terms before it is criticised.
Having understood it on its own terms, the criticism is this: the appeal to inscrutability is not a theodicy at all. It is the withdrawal of the claim to theodicy. To say “God’s plan is beyond human understanding” is to abandon the project of showing that God’s plan is good, and therefore to abandon the project of reconciling divine attributes with observed suffering. If God’s goodness is of a kind so radically different from human goodness that it produces outcomes indistinguishable from the operation of a universe containing no good God at all, then the word “good” as applied to God has been stripped of its ordinary meaning. We can no longer say “God is good” in any sense that connects to the moral vocabulary we use in every other context. The word becomes a ritual affirmation detached from any content that could be tested, evaluated, or even coherently described.
This matters enormously, because the claim of divine goodness is not merely a metaphysical curiosity. It is the foundation of the moral authority that religious institutions claim to exercise, the justification for moral edicts, for the authority of scripture, for the demand that believers subordinate their own moral judgement to divine command. If “God is good” means only “God’s ways are inscrutable,” then the entire moral architecture built on divine goodness collapses. You cannot derive a binding moral code from an inscrutably alien will. You cannot ask people to trust a being whose defining characteristic is that its purposes cannot be understood or evaluated by the people required to defer to them. A God whose goodness is demonstrable only to himself, and whose instructions must therefore be accepted on faith alone, is a God whose authority rests on nothing that moral reasoning can reach or verify.
The Book of Job is, in this reading, not a satisfying theodicy but an honest confession that no satisfying theodicy exists. God’s answer to Job is not an explanation; it is a display of power. Job is silenced, not persuaded. And the moral of that silencing, which many readers have drawn across the centuries, is that questioning divine authority is presumptuous, rather than that divine goodness has been vindicated by any argument or evidence. That is not a moral lesson with any claim to our respect. It is a capitulation to power dressed as wisdom, and the dressing should not deceive us.
There is a further dimension worth naming. The inscrutability defence, deployed consistently, would prevent theology from making any positive claim about God’s character at all. If God’s goodness is beyond human comprehension, then so is his justice, his mercy, his love, and his anger. All of the attributes that make the God of the major monotheisms personally significant, morally authoritative, and worthy of worship are, on the inscrutability view, beyond human evaluation. A being whose nature is wholly inscrutable to human understanding is a being about whom human beings can say nothing substantive whatsoever. That is not the God of any living religious tradition. It is a placeholder, a linguistic convenience, a name without content. The inscrutability defence, pushed to its logical conclusion, dissolves the very God it was invoked to defend.
Gratuitous Suffering and the Test Case Every Theodicy Fails
Philosophers of religion have developed a concept specifically designed to test whether any theodicy is adequate: the concept of gratuitous evil, meaning suffering that serves no compensating good, suffering that a genuinely omnipotent being could have prevented without cost to any value that would justify allowing it. The question is not whether suffering in general can be reconciled with divine goodness, but whether there exist specific instances of suffering so extreme, so purposeless, and so devoid of compensating benefit that no theodicy can account for them. The claim, sharpened by philosophers such as William Rowe, is that such instances obviously exist, and that their existence constitutes strong evidence against the existence of an omnipotent, omnibenevolent God.
Consider the specific case that Rowe uses in his canonical formulation: a fawn trapped in a forest fire, dying slowly over days from its burns, in agony, with no human being present to witness it and no mechanism by which its suffering could plausibly contribute to any good detectable by any moral accounting system we possess. The free will defence is entirely inapplicable: no human free will is involved in any step of the process. The soul-making defence is inapplicable: the fawn does not develop moral character through its ordeal. The greater good defence requires us to identify what good is served by this particular instance of suffering, and no candidate good is available, nor has any theologian supplied one. The inscrutability defence simply concedes the point: we cannot say why this suffering occurs, only that there must be a reason we cannot see. But “there must be a reason” is not an argument; it is the conclusion assumed within the premise, which is precisely the circular reasoning the argument should be designed to avoid.
Now scale up from the fawn. Childhood cancer is not caused by free will. It is not a mechanism of soul-making for an infant too young to process the experience morally or spiritually. It does not serve any greater good that has ever been identified or even plausibly suggested. It is not, in any meaningful sense, a divine lesson, because the child dying from it is not in a position to learn anything from the suffering, and the suffering involved is grotesquely disproportionate to any educational purpose it might conceivably serve. Every theodicy either fails to account for childhood cancer specifically, or retreats to inscrutability, which is not an account at all but an admission that no account is available and that the absence of one should not count against the theological conclusion. That is special pleading of the most transparent kind, and it should be named as such.
This is the test that theology consistently fails, not on some abstract philosophical level, but in the most concrete, human, and morally serious terms available. The problem of evil is not a parlour game for sceptical philosophers. It is the question that every parent sitting beside a dying child is implicitly asking when they demand to know why this is happening, and what the theology of “everything happens for a reason” is prepared to offer as an answer. What it offers is one of the four defences examined in this essay. None of them survives the encounter with the specific, irreducible fact of the child dying in the hospital bed beside people who love her and can do nothing to prevent it. The gap between what the theology promises and what it delivers, when pressed on this specific terrain, is not a gap that any additional theological ingenuity can close, because the problem is not a lack of ingenuity but a lack of evidence.
The Moral Cost of the Comfortable Answer
The intellectual failure of theodicy would be significant enough on its own. But the failure has consequences that are moral rather than merely philosophical, and those consequences deserve to be named directly, because they fall on real people in identifiable ways.
The first consequence is the instrumentalisation of the victim. To say that a child’s death serves a divine purpose is to convert the child from an end in herself into a means toward someone else’s end. This is the precise moral move that Kant identified as the fundamental violation of human dignity: treating a person not as an end in herself but as a tool for the satisfaction of some external purpose. The fact that the external purpose is divine rather than human does not alter the moral character of the move; if anything, it amplifies the offence, because the child has no capacity to consent to being used as an instrument of divine pedagogy, and the God who uses her in this way is, by hypothesis, fully aware of the suffering involved and capable of achieving any educationally or morally good end by other means, since his power is unlimited. The theodicy, in making suffering purposive, makes the one who suffers instrumental. That is not a defence of divine goodness. It is a description of divine exploitation dressed in the language of care.
The second consequence is the silencing of legitimate moral outrage. If suffering is divinely ordered and serves a purpose, then the appropriate response to it is acceptance, trust, and the subordination of human moral judgement to divine will. But moral progress in human history has been achieved, consistently and without exception, by people who refused to accept suffering as the natural or divinely ordained order of things, who refused to accept that poverty, disease, oppression, and cruelty were part of a plan that should be trusted rather than a problem that should be solved. Theodicy, insofar as it counsels acceptance of suffering as divinely willed, functions as a brake on moral progress. It did so in the long resistance to anaesthesia, argued by some theologians to circumvent the divinely ordered pain of childbirth. It did so in theological justifications of slavery, which invoked the curse of Ham and the Christianising mission to give scriptural warrant to a criminal institution. It did so in the opposition to vaccination, to the germ theory of disease, to countless medical and social advances that involved the human decision to refuse to accept suffering as though it were a gift from a benevolent architect. The consequences of theodicy are not confined to the philosophical seminar room. They are written in the history of preventable deaths, and that history is long.
The third consequence is the moral exculpation of genuine wrongdoing. If suffering is divinely ordered and everything happens for a reason, then abusers, torturers, and genocidal regimes are, on some level, instruments of divine purpose. This conclusion is not usually drawn explicitly, because it is too obviously monstrous, but it follows from the logic of theodicy with uncomfortable directness. The theodicy that says the Holocaust served a divine purpose must say, if it is to be consistent, that those who executed it were serving that purpose. The theology that says every event in a rape survivor’s life, including the rape itself, was part of God’s plan for her, is a theology that has made itself morally indistinguishable from a defence of the rapist’s role in that plan. These conclusions are not drawn by the theologians who construct the defences, but they are entailed by them, and the fact that they are not drawn is evidence that the theologians are relying on moral intuitions that directly contradict their own theodicies. The intuitions are sound. It is the theodicies that should be abandoned.
You can find a fuller examination of the specific moral consequences of the “God’s plan” framework in the earlier piece on this site, God’s Plan and the Problem of Moral Cowardice, which traces the way this theological framework has been deployed to suppress legitimate responses to institutional failure.
The Epistemological Problem: Why Theodicy Cannot Even Get Started
There is a level at which the argument against theodicy is more fundamental than any of the specific objections raised above, because it concerns not what the theodicies say but what they would need to establish before saying anything at all. Every theodicy begins by assuming the existence and attributes of the God whose ways it is attempting to justify. But the existence of an omnipotent, omnibenevolent, omniscient God is precisely what the problem of evil places under doubt. The theodicy is therefore being offered as a defence of a conclusion that has not yet been established, against evidence that suggests the conclusion is false. Framed this way, the entire enterprise of theodicy is seen to be question-begging from its first move.
This is not a minor methodological quibble. It goes to the heart of what theodicy is and what it can legitimately claim to do. If the existence and goodness of God were established independently of the argument from evil, then theodicy would be a reasonable project: given that we know God is good and omnipotent, how do we account for suffering? But God’s existence and goodness are not established independently of the argument from evil. The argument from evil is one of the primary challenges to belief in God, and the attempt to answer it by assuming the attributes it is designed to challenge is a circular proceeding that establishes nothing beyond what was already assumed at the outset.
This is also why the appeal to inscrutability is so epistemologically peculiar. The argument runs: we know God is good, and therefore the suffering we observe must serve good ends that we cannot perceive. But we know God is good only if we have some independent evidence that God is good. If the suffering we observe is consistent with a God who is indifferent, or malevolent, or simply nonexistent, and if we have no independent evidence of God’s goodness that is not derived from the very theological tradition whose credibility the suffering is calling into question, then the inscrutability defence is simply assuming what it needs to prove. The logical structure is identical to that of a defendant claiming that the charges against him must be false because he is, by nature, a good person, and offering as evidence of his good nature only his own assertion of it, unsupported by anything external. Courts do not accept that reasoning, and neither should anyone engaged in an honest assessment of the evidence.
The methodological standard that should apply here is not specially demanding or philosophically unusual. It is the same standard that applies to any empirical claim about the world. A habit of demanding evidence proportionate to the magnitude of a claim, and of declining to grant certainty beyond what the evidence warrants, is not scepticism in any hostile or nihilistic sense; it is simply the ordinary intellectual hygiene that distinguishes reliable knowledge from wishful thinking. Theodicy, measured against that standard, begins with an unfounded assumption and proceeds from it by means that are neither empirical nor logically demonstrative. It is a philosophical project that cannot establish its own starting point, and whose conclusions therefore float free of any contact with the evidence it claims to address.
There is an additional epistemological problem specific to the greater good and inscrutability defences. Both defences rely, at crucial moments, on what philosophers call “sceptical theism,” the view that we should not expect to be able to discern God’s reasons for permitting suffering, because our cognitive capacities are too limited relative to the scope of divine purposes. This is a coherent position, but it comes at a significant cost that sceptical theists rarely acknowledge fully. If we cannot trust our moral perceptions to track what an omniscient God would regard as good or evil, then we equally cannot trust the moral perceptions that led us to regard God as good in the first place, or the moral intuitions that underlie any reading of scripture as divinely inspired, or the moral sense that allows believers to distinguish genuine religious experience from delusion. Sceptical theism, consistently applied, undermines not only the argument from evil but every positive ground for belief in a good God. The defence defeats its own purpose by sawing off the branch on which the entire theological enterprise is seated.
What Theodicy Reveals About the Theology It Defends
Theodicy, examined carefully, does not merely fail to solve the problem it addresses. It reveals something about the theology it is designed to defend, and what it reveals is not reassuring to that theology.
Consider what happens to the concept of divine omnipotence as theodicy proceeds through its various forms. The free will defence implicitly constrains what God can do: he cannot create free beings who reliably choose good, or at least not without violating some principle of libertarian freedom that the theologian treats as sacrosanct, without ever explaining why this principle should constrain an omnipotent creator. The soul-making defence similarly constrains God: he cannot produce morally developed souls without the resistance that suffering provides, which means his creative options are limited by the requirements of a pedagogical process that he did not apparently design with any particular efficiency in mind. The greater good defence imports a framework of consequences within which God must operate, as though he were a consequentialist agent constrained by a moral calculus that he did not himself determine and cannot apparently override. In each case, omnipotence is quietly eroded. God becomes a being of very great power who is nonetheless constrained, in various ways and for various reasons, from preventing the suffering that the theodicy is supposed to account for. The God who emerges from this process is not the God of classical theism. He is a significantly diminished deity whose limitations happen, with suspicious convenience, to coincide with whatever constraint is needed to make the current theodicy cohere.
Similarly, consider what happens to the concept of divine goodness as theodicy proceeds. The soul-making defence implies that God’s commitment to a particular developmental process outweighs his direct concern for the welfare of those being developed, including those who die before the process can bear any fruit. The greater good defence implies that God’s pursuit of outcomes justifies means that we would condemn without hesitation in any human agent who employed them. The inscrutability defence implies that “divine goodness” is a term so radically disconnected from ordinary moral language that it cannot be used to generate any expectation about what a good God would or would not allow, which renders it a term without practical content. In each case, what is called “divine goodness” becomes progressively less recognisable as goodness in any sense that connects to the moral concepts we use in the rest of our lives and in all our other moral reasoning.
The God who emerges from theodicy, if we take the defences seriously, is not the God of the pulpit or the prayer book: loving, personal, intimately concerned with the welfare of each individual soul. He is something more like a cosmic process with purposes that are opaque to the beings caught up in them, operating through mechanisms that include the prolonged agony of children and the mass death of innocents, toward ends that cannot be specified in advance and cannot be evaluated after the fact because they are, by definition, invisible to any human moral assessment. This is not a source of comfort. It is, if such a being exists at all, something to be endured rather than loved. And the theology that has produced him in the course of trying to defend his goodness has, in so doing, made one of the more powerful arguments for the secular conclusion that the problem of evil is not a problem to be solved by theodicy but a problem to be acknowledged honestly, and that the acknowledgement leads, by the most direct route available, to the conclusion that the traditional God of theism does not exist.
Religion’s Relationship with Tragedy: The Institutional Harvest
There is one further dimension of the “everything happens for a reason” framework that deserves examination: its function for religious institutions rather than for individual believers. Because while individual believers may reach for the phrase in moments of genuine grief, with genuine compassionate intent and no ulterior motive, the theological apparatus that underwrites it serves institutional interests that are worth making explicit.
Tragedy is, as the companion piece on this site observes, among religion’s most reliable recruiting grounds. The connection between grief and religious conversion is well-documented: people in the aftermath of bereavement, illness, or catastrophe are more susceptible to religious appeals than people in stable, comfortable circumstances. This is not a cynical observation about religious institutions, most of which contain people of genuine compassion and sincerity; it is a structural observation about the conditions under which theological frameworks become psychologically compelling. Grief destabilises the cognitive frameworks through which we ordinarily navigate the world, creates an urgent demand for meaning, and makes the person experiencing it vulnerable to systems that offer meaning in exchange for assent to claims that would receive more sceptical scrutiny in calmer times.
“Everything happens for a reason” is, in this context, not merely a consolation; it is also a conversion mechanism. It offers meaning at the moment when meaning is most desperately sought, and it attaches that meaning to a theological framework that requires ongoing participation in a religious community to sustain. The person who accepts the framework in the hospital waiting room is thereafter embedded in a web of commitments, practices, and beliefs that extend far beyond the original moment of vulnerability. The theodicy does not merely comfort; it recruits. And the fact that the comfort is genuine for many people does not alter the fact that the theological claims that underwrite it do not withstand scrutiny, or that the institutional benefits to religion of offering that comfort in that moment are substantial enough that they deserve to be named as part of the full picture.
None of this constitutes an argument against compassion, against community, or against the genuine goods that religious communities often provide to people in crisis. Those goods are real and deserve acknowledgement. The argument is specifically and only against the theological claim: the assertion that suffering has a divine purpose, that everything happens for a reason, that God’s plan encompasses and justifies the death of children and the drowning of 227,000 people in a single morning. That claim is false, in the sense that no evidence supports it, that every argument constructed to defend it either fails when applied to specific cases or quietly abandons the divine attributes it is supposed to vindicate, and that the moral cost of accepting it is borne by the most vulnerable people at the most vulnerable moments of their lives, who deserve honesty rather than a consolation that cannot be cashed out.
Religious communities, it should be acknowledged, also contain thinkers who have been deeply troubled by this problem and who have refused the easy consolations. Dostoevsky’s Ivan Karamazov does not speak as an atheist polemicist; he speaks as someone who has taken theodicy seriously enough to reject it on moral grounds, even while remaining unable to fully resolve the question of God. The tradition of protest theology, of wrestling with divine injustice, of saying plainly that the suffering of the innocent cannot be reconciled with an omnipotent and wholly good God, is a minority tradition within religious thought, but it is an honest one, and it reaches, in its most rigorous forms, conclusions that are not very distant from the secular conclusion that the God of classical theism does not exist. The problem of evil is not a weapon manufactured by atheists. It is a recognition that theology’s own central claims are in tension with the evidence, a recognition that honest theologians have shared, and that the defences mounted in response have not resolved.
Facing the Evidence Without the Consolation
The secular alternative to theodicy is not comfortable, and it would be dishonest to pretend otherwise. To say that suffering has no divine purpose, that a child’s death serves no plan, that the tsunami was a geological event with no moral meaning, is to remove a framework that many people find genuinely sustaining and to replace it with the acknowledgement that the universe is indifferent and that some suffering is simply without justification, without purpose, and without recompense in any life to come. This is a harder position to inhabit than the one theodicy offers. It makes no promises about the next life, offers no narrative in which the dead child’s suffering is retrospectively vindicated, and demands that the living face grief without the scaffolding of divine purpose to lean upon.
But the question is not which view is more comforting. The question is which view is true, and on that question, Carl Sagan’s counsel remains both honest and genuinely humane: “The world is so exquisite, with so much love and moral depth, that there is no reason to deceive ourselves with pretty stories for which there’s little good evidence. Far better, it seems to me, in our vulnerability, is to look Death in the eye and to be grateful every day for the brief but magnificent opportunity that life provides.” This is not stoicism in the bleak, austere sense. It is a recognition that the honest confrontation with mortality and suffering, without the analgesia of theodicy, is not only philosophically superior but humanly richer, because it is not built on a consolation that requires us to pretend that a child’s death was a good idea, or that the agony of the burned fawn served some cosmic purpose that we simply lack the perspective to appreciate.
The secular response to suffering is not that suffering is fine or that it serves a purpose. It is that suffering is precisely not fine, that it demands a response rather than acceptance, that the appropriate reaction to a child with leukaemia is not to trust the divine plan but to fund cancer research, to sit with the parents in their grief, to demand better healthcare systems, and to refuse to accept that the suffering was necessary. This response is fully available to the person who has abandoned theodicy. It is not available, without serious internal contradiction, to the person who believes the suffering was divinely ordered, because you cannot simultaneously accept that an event was willed by an omnipotent and perfectly good God and also campaign to prevent similar events in the future. If God willed it as part of a good plan, then preventing its future occurrence would thwart that plan. The theodicy, if taken seriously as a guide to action, counsels acceptance. The secular view counsels resistance. The moral superiority of resistance over acceptance, in the face of preventable suffering, is not a difficult case to make, and the history of secular humanism is, in large part, the record of what that resistance has achieved.
What secular humanism offers in place of the consolation of theodicy is something more demanding and, in the long run, more honest: the recognition that we are the ones responsible for reducing suffering, that no divine plan will do this work on our behalf, that the bereaved deserve our presence and our practical assistance rather than our theological assurances, and that meaning, where it is to be found at all, is constructed by human beings through their relationships and their choices rather than discovered as the pre-existing content of a divine agenda. This is not a cold or impoverished vision of human life. It places the weight of responsibility squarely on human shoulders, and the history of medicine, of public health, of the abolition of slavery, and of the expansion of human rights suggests that human shoulders are capable of bearing it, when they are not first convinced that the bearing is unnecessary because God has a plan.
The honest secular account of suffering also has the significant advantage of being consistent with what we actually observe. The distribution of suffering in this world is not correlated with moral desert. Good people suffer horribly; cruel people sometimes flourish. Natural disasters strike without reference to the virtue of their victims. Disease does not discriminate between the devout and the sceptical. The randomness of suffering, its complete indifference to the moral qualities of those it afflicts, is precisely what we would expect if the universe operates without divine supervision, and precisely what theodicy must explain away by means of increasingly implausible appeals to invisible compensating goods or inscrutable divine purposes. The secular account requires no such supplementation. It matches the evidence directly, without the intermediate machinery of a plan that cannot be identified, evaluated, or falsified. That is not a small philosophical advantage. It is the difference between a theory that fits the data and one that has to keep modifying its auxiliary hypotheses to avoid being refuted by it.
The Verdict
Theodicy, across its major forms, is a project that has failed on its own terms. The free will defence cannot account for natural evil and cannot explain why an omnipotent God did not actualise a better world from within the vast range of possible worlds available to him. The soul-making defence converts individuals into instruments of a divine curriculum they cannot consent to, and cannot account for those who die before the curriculum can produce any of its supposed benefits. The greater good defence, in claiming that all suffering serves invisible purposes, produces a concept of divine goodness so elastic that it is consistent with any possible state of affairs and therefore explains nothing at all. The appeal to inscrutability concedes the failure of all the others and substitutes trust for argument, while simultaneously withdrawing the content from the very attributes that the theodicy was constructed to defend.
What remains, once the defences are cleared away, is the problem as Epicurus stated it twenty-five centuries ago and as it still stands today: an omnipotent, omnibenevolent, omniscient God is incompatible with the world we observe. The incompatibility is not a puzzle awaiting a sufficiently clever theological solution. It is evidence, consistent and cumulatively compelling, that the God described by the traditions that generated this problem either does not exist, or is not what those traditions have claimed him to be. Both possibilities lead away from theodicy and toward the secular conclusion that the universe operates without reference to human suffering, that this fact places a moral demand on us rather than removing one, and that the demand is to act rather than to accept.
The moral cost of pretending otherwise is paid by the bereaved, by the dying, and by everyone whose legitimate moral response to suffering is redirected from action into acceptance by a theology that has learned to dress its failures in the language of mystery and plan. “Everything happens for a reason” is not wisdom. It is the sound of theology reaching for comfort it cannot honestly provide, and declining to acknowledge that the reach is in vain. The honest answer, the one that respects the dead and the dying rather than converting them into theological furniture, is that some things happen for no good reason at all, that this is a moral outrage rather than a divine arrangement, and that the appropriate response is not to trust the plan but to reject, clearly and without equivocation, the claim that any such plan exists.
That rejection is not despair. It is the beginning of a moral seriousness that theodicy, for all its centuries of effort and for all the genuine intelligence that has been brought to bear upon it, has never managed to achieve.
References
Epicurus, fragments preserved in Lactantius, De Ira Dei, c. 313 CE.
John Hick, Evil and the God of Love, 1966.
Alvin Plantinga, God, Freedom, and Evil, 1974.
William Rowe, “The Problem of Evil and Some Varieties of Atheism,” American Philosophical Quarterly, 1979.
John Stuart Mill, Autobiography, 1873.
Steven Pinker, Enlightenment Now, 2018.
Carl Sagan, Billions and Billions, 1997.