Three claims get merged in almost every argument from evil. "I do not see God." "I cannot explain why God permits this." "There is no God." The first two are reports about the speaker. Only the third is a claim about reality, and only the third carries a burden of proof.

This article follows a recent exchange with an atheist and the replies it drew. It sorts the arguments into what each one establishes and what it does not.

Two kinds of evil

Murder, genocide, and torture are moral evil. They come from human choices. Earthquakes, childhood cancer, and animal pain are natural evil. They come from no one's choice. An answer to one does not answer the other, so each gets its own argument below.

The argument from evil is an argument about ought

The skeptic's case ran like this: evil exists, a good God would not permit it, therefore there is no God. The second sentence carries the whole argument, and the skeptic has to defend it.

He asked why God did nothing about Rwanda and Bosnia. "Did nothing" means "should have done something." That is a moral claim. It charges God with a failure, so a standard of right and wrong has to stand over the act and over the accused.

If nature is all there is, where does that standard live? Hume held that no pile of "is" statements yields an "ought." Chemistry, instinct, and habit explain what people do and why they recoil. They do not show that the killers were wrong rather than merely effective.

Some atheist philosophers defend objective morality without God. The question is whether a world of matter alone supplies the authority they claim.

C. S. Lewis told this story about himself. He said his argument against God was that the universe seemed cruel and unjust. Then he asked where he had gotten his idea of justice.

A broken arm is damage. A broken promise is a debt. The arm can be measured. The debt cannot be weighed, and people still collect it. If ten million people approved of torturing a child, it would still be wrong.

The strongest version is evidential

William Rowe's influential 1979 argument is evidential. He held that no one has shown evil and God to be logically inconsistent, and argued instead that the suffering we observe makes God's existence less probable.

That is a probability claim. It compares how likely the observed suffering is if God exists with how likely it is if God does not, and it also depends on how likely God was to begin with. Both are judgments. The skeptic has to defend them.

Allowed is not authored

I hold that God made humans able to choose and answerable for the choice. A being who can choose good can choose evil. Removing the second removes the first.

The skeptic treats "God allowed it" as if it meant "God did it." A judge who lets a man stand trial did not commit the crime. A parent who lets a grown son ruin himself did not make the choice. The killer is responsible. So is the crowd that looked away.

If God should have overridden those choices, that case has to be argued. Sliding from "allowed" to "authored" is not an argument.

Two answers that did not work

One commenter raised Cain. Genesis 4:7 has God warning Cain that sin is crouching at the door and that he must master it. After the murder, God holds Cain answerable (Genesis 4:9–10). This happens before Sinai and before any commandment is numbered. Cain is judged by a standard that stood over him without a written code. That supports the argument above.

The same commenter said sharks and wolves do not kill their own kind. That is false. L. David Mech documented territorial gray wolves killed by other wolves. Craig Packer and Anne E. Pusey documented infanticide by incoming male lions. R. Grant Gilmore, Jon W. Dodrill, and Patricia A. Linley documented sand tiger shark embryos consuming other intrauterine embryos and ova.

The moral point fails even if the facts were right. Animal behavior describes what animals do. It does not say what anyone ought to do. No court has ever tried a wolf. We try people, and we do it because we hold people to a moral law.

He also proposed that God is a process, so God and evolution are one. That is a redefinition. It supplies no evidence that the process is God or that God is the process. It also removes the original charge. A process cannot be blamed for doing nothing about Rwanda. A person can keep the definition or the charge, but not both.

Where the skeptic's tools stop

The skeptic asked why no telescope has seen the supernatural. A telescope is built to detect physical objects. It has not seen a thought, a duty, a number, or another person's mind either. We know those by what they do. The point is about fit between instrument and question, not that every kind of knowledge works the same way. Jesus used the wind: you hear it, you see what it does, you never see it (John 3:8).

If the only instrument allowed is one built for physical objects, and an empty result is announced as a verdict, the rule was set before the test.

Two atheist writers have spoken about motive. Thomas Nagel has written that he wants atheism to be true and hopes there is no God. Richard Lewontin wrote that materialism is absolute because "we cannot allow a Divine Foot in the door." Motive does not refute a conclusion. It shows that evidence may not be the only thing setting the rule.

Nagel's Mind and Cosmos argues that consciousness, reason, and value are not adequately explained by physics and chemistry. He did not become a believer. He treated the problem as serious and unresolved by materialism.

Darwin raised a related doubt in an 1881 letter to William Graham. He questioned whether the convictions of a mind developed from lower animals can be trusted, and he set that doubt against his own sense that the universe was designed. Darwin did not use it to argue for theism. I apply the same question to the conviction that nature is all there is.

So the questions for the skeptic stand. What is the positive evidence that nothing exists beyond nature? What would change his mind? If the answer is nothing, the door was locked before the test.

An argument inside the Christian house

A Christian reader raised a fair point. In his view, human responsibility does not require the ability to choose otherwise. It requires acting voluntarily from one's own desires. This is compatibilism. He cited Acts 4:27–28, where Herod, Pilate, and the others did what God's hand and plan had predestined.

I accept what that passage states. The crucifixion fell within God's plan, and the men who carried it out remain accountable. Genesis 50:20 is a related case: the brothers meant evil, and God meant good.

Christians disagree about how sovereignty and human choice fit together. These texts do not settle it. Acts 14:16 has God allowing the nations to walk in their own ways, so Scripture speaks of permission as well as decree.

My argument to the skeptic needs only a narrower claim. God's permitting an act does not, by itself, establish that He authored the evil of the act, and the skeptic must demonstrate that connection rather than assume it.

Comfort for a grieving parent is real. Comfort alone does not prove a theological claim.

Natural evil is a separate argument

A storm is not a murder. The argument from natural evil has to be made and answered under its own name.

"Allowed, not authored" does not carry this one. Isaiah 45:7 shows why. The King James Version has God creating "evil." The Hebrew word is raʿ, set against shalom, and the ESV, NASB, NIV, and NKJV render it "calamity" or "disaster." The context concerns Cyrus and God's purposes for the nations, not the creation of moral wickedness. That answers a lexical objection. It does not answer natural evil. If God creates calamity, God is the author of the calamity. Romans 8:20 says creation was subjected to futility "by the will of the one who subjected it."

The Christian answer has to rest on something else: the Fall, divine judgment, or a reason God has not disclosed. A claim like "sin is the ultimate cause of natural disasters" is a theological premise. A skeptic will point to suffering before humans existed and reject it. It has to be defended, not asserted.

The grounding question returns here. A tsunami is a physical event. The evil is the suffering it causes, and calling that suffering evil is a moral judgment. At most, natural evil lowers God's probability, and how far depends on estimates the skeptic must defend.

What each claim establishes

Humans commit moral evil. Established by observation.

Genesis holds Cain accountable before Sinai. Supported by the text.

Acts 4:27–28 places the crucifixion in God's plan and keeps the actors guilty. Supported by the text.

These texts settle how sovereignty and human choice fit together. Not established.

Evil makes God logically impossible. Not established; Rowe himself grants this.

Evil makes God less likely. Depends on estimates the skeptic must defend.

Isaiah 45:7 says God creates moral wickedness. Not supported by the Hebrew or the context.

Every natural evil has a known divine purpose. Not established.

Nature alone grounds objective moral obligation. Contested; the burden is on the naturalist.

What the argument establishes

Evil does not establish that God does not exist. The logical problem asks whether God and evil can both be real. Rowe himself says no one has established that they are logically inconsistent. The evidential problem asks whether the suffering we observe makes God less likely, and that depends on estimates the skeptic has not defended.

"I do not see God" is a report about the speaker. "I cannot explain why God permits this" names a hard question. Only "there is no God" is a claim about reality, and it needs evidence and a stated condition under which the speaker would change his mind.

The question is not whether suffering is real. It is whether suffering establishes the conclusion the skeptic draws from it.

© 2026 Charles Mason

Sources

William L. Rowe, "The Problem of Evil and Some Varieties of Atheism," American Philosophical Quarterly 16, no. 4 (1979): 335–341.

Charles Darwin, "Letter to William Graham, July 3, 1881," Darwin Correspondence Project, letter no. 13230.

Thomas Nagel, The Last Word (Oxford University Press, 1997), 130–131. Publisher's book page.

Thomas Nagel, Mind and Cosmos: Why the Materialist Neo-Darwinian Conception of Nature Is Almost Certainly False (Oxford University Press, 2012).

Richard C. Lewontin, "Billions and Billions of Demons," The New York Review of Books, January 9, 1997.

L. David Mech, "Buffer Zones of Territories of Gray Wolves as Regions of Intraspecific Strife," Journal of Mammalogy 75, no. 1 (1994): 199–202.

Craig Packer and Anne E. Pusey, "Adaptations of Female Lions to Infanticide by Incoming Males," The American Naturalist 121, no. 5 (1983): 716–728.

R. Grant Gilmore, Jon W. Dodrill, and Patricia A. Linley, "Reproduction and Embryonic Development of the Sand Tiger Shark," Fishery Bulletin 81, no. 2 (1983): 201–226.

C. S. Lewis, Mere Christianity, Book II, chapter 1.