What Would It Take to Believe a Miracle?

Hume and the Problem of Miracles

David Hume was a Scottish philosopher who came up with something that really stuck around. In An Enquiry Concerning Human Understanding, he made an argument about miracle claims that’s still the best response. It’s simple, it’s elegant, and nobody’s really refuted it.

What Hume Noticed

Start with what a miracle is: by definition, it breaks a natural law. Hume makes a simple observation: our confidence in natural laws comes from what we’ve seen happen over and over. We’ve watched countless people die. None of them came back to life. We’ve seen many people sick. Prayer alone hasn’t healed them. We’ve observed the stars and planets. None of them stopped moving when someone asked.

So how much should we believe in natural laws? It should match how much evidence supports them. Gravity has trillions of observations behind it. The fact that dead things stay dead? Billions of observations. These aren’t guesses or tentative ideas. They’re confirmed far beyond doubt.

Now imagine someone says a miracle happened. They claim someone died and came back, or the sun stopped moving, or an amputated hand grew back. They’re saying one of our best-supported laws was broken. And their proof? Someone’s account of what they saw.

Here’s Hume’s key insight: you have to weigh testimony against the law it’s contradicting. Miracle claims rest on someone’s account that violates a law backed by the entire history of human observation. What’s more likely: that the law was broken, or that the person was mistaken, exaggerating, lying, or misinterpreting what they saw?

Hume’s conclusion: testimony can’t prove a miracle happened. The testimony would need to be incredibly strong, backed up by so many sources, so reliable, so thoroughly documented, that it would be miraculous if the testimony were false. No testimony in history has been that strong.

Why This Actually Matters

Miracles are what religions use as their main evidence. Christianity hangs on resurrection. Islam points to Muhammad splitting the moon. Hinduism points to statues drinking milk. Every religion relies on miracle claims.

Hume’s argument shows the problem: testimony is the only evidence we have for miracles. And no matter how many people swear they saw something, no matter how honestly they believe it, no matter how detailed the story is: testimony can’t compete with the overwhelming evidence that natural laws hold.

Applied Across Religions

Christian Miracles: The Resurrection

Christian theology says Jesus was executed, stayed dead for three days, then came back to life. That breaks the law that dead humans stay dead. We’ve never once seen someone spontaneously resurrect. Ever. That law has the entire history of the human race behind it.

The evidence? The Gospels: written decades after the event by people with theological reasons to believe it. And the Gospels contradict each other about the details. Who went to the tomb, what they found there, where Jesus appeared, and to whom: the four accounts tell it four different ways. Paul, the earliest Christian writer of all, describes appearances but never mentions an empty tomb.

These accounts don’t match. They were written by believers, not impartial observers. They describe events no historian at the time recorded, even though supposedly people came back from the dead all over Jerusalem. Rome had good historians. So did the Jewish community. Nobody mentioned it.

By Hume’s thinking, this testimony isn’t strong enough. What’s more likely: that dead people came back to life for the first time in history, or that grief-stricken disciples had visions, and the story got bigger as it was retold over decades? The second fits what we know about psychology and how stories change. The testimony doesn’t beat the enormous evidence that dead humans stay dead.

Islamic Miracles: The Moon Split

Islamic tradition says Muhammad split the moon in half: a miracle everyone could see. That would break natural law. The moon is massive, governed by gravity and rotation. Splitting it would violate physics and be visible everywhere on Earth.

But no astronomy from that time recorded this. No crater marks where it supposedly fused back. No geological evidence. Nothing from non-Muslim historians. Only Islamic texts written centuries after Muhammad died.

By Hume’s measure, this evidence is weak. What’s more likely: that the moon was split and immediately re-formed leaving no trace, or that this became a story in the tradition over centuries? The second fits how religions develop. The testimony can’t overcome the total absence of physical evidence and contemporary records.

Hindu Miracles: Ganesha Drinking Milk

In 1995, Hindu tradition claimed statues of Ganesha drank milk offered in temples across India and worldwide. Thousands said they watched statues drink milk through their trunks.

But nobody did chemical testing. The milk supposedly absorbed by statues vanished. Nobody explained how a stone or metal statue could consume liquid. No lab ever recreated it.

Scientists found a simple explanation: capillary action. Porous stone wicks liquid across its surface, and it looks like the statue is drinking. A natural explanation fits the observations perfectly, no miracle needed.

By Hume’s standard, witnesses seeing something ambiguous isn’t enough to prove a miracle when a natural explanation works. The testimony can’t compete with the law that statues don’t consume things.

Ancient Greek Miracles: Asclepius Healings

Ancient Greeks said Asclepius, the god of medicine, healed people miraculously in temples. Pilgrims came, slept in the temple, and reported healing visions.

The evidence: old texts and inscriptions. But no medical records. No follow-up to see if people actually stayed healed. The accounts came from believers, not skeptical observers. The healings were described as spiritual experiences, not medical facts.

Modern psychology explains this: belief in treatment creates real physical changes. Stress-related illness improves when stress drops. Pain feels better when you’re confident. Healing through belief is natural. It doesn’t need a god.

By Hume’s measure, believers describing unclear phenomena can’t prove a miracle when natural explanations work. The placebo effect, stress-related illness improving, selective memory, these explain what happened. The testimony isn’t strong enough.

How People Respond to Hume

Response 1: “Hume assumed natural laws can’t be broken.”

He didn’t. He based it on evidence. We believe in natural laws because we keep experiencing them. Gravity works everywhere. Things that die stay dead. That experience gives us good reason to trust natural law. Hume’s point is simple: testimony alone can’t outweigh that kind of experience.

Response 2: “Hume’s argument is circular: he assumes miracles can’t happen.”

That’s not right. Hume doesn’t assume miracles are impossible. He points out that bigger claims need bigger evidence. Testimony is only “big evidence” if it would be miraculous for it to be false. No testimony has ever been that strong. Until one is, Hume’s argument works.

Response 3: “Eyewitness testimony should be enough.”

But it’s not reliable. Psychology shows eyewitness testimony gets things wrong a lot, especially when people are emotional or surprised. Someone seeing something weird might decide it was miraculous because they expected it to be.

And ancient eyewitness testimony? Even weaker. The Gospels were written decades after events by people we don’t know. We don’t know if they saw things themselves. We don’t know if they were copying earlier sources or making things up. We can’t question them. Ancient accounts don’t have enough corroboration or detail to beat the enormous evidence for natural law. Plus we know the observers knew nothing of science. They literally didn’t know that air exists, much less what it was made of.

Response 4: “God could do a miracle so obvious everyone would see it.”

Sure. God could do something undeniable observed by everyone on Earth at once. But that hasn’t happened. Every miracle claim in history is unclear, argued about, poorly documented, and explainable by natural causes. If God wanted testimony to prove a miracle, God could make it obvious. The fact that no such obvious miracle exists suggests miracles haven’t actually occurred.

Why Testimony Doesn’t Work for Miracles

Testimony has problems when it comes to proving miracles:

  • People interpret through their beliefs: When something’s ambiguous, believers see what they expect to see
  • Ancient accounts are old: The further we are from the event, the more it changes in retelling
  • Believers have reasons to believe: Religious people are motivated to see miracles
  • Testimony contradicts itself: We can spot where stories don’t match up
  • Every religion has testimony: But they all support different, opposite gods

That last one matters. Muslims point to Quranic miracles. Hindus point to deity miracles. Buddhists describe enlightenment miracles. Christians describe resurrection. Each tradition’s evidence supports a different god. They can’t all be true. But the testimony quality is similar in each.

If testimony could prove miracles, you’d have to believe in all these contradictory gods. But nobody does. We’re skeptical of testimony from other religions and trusting of our own. That reveals something: we don’t actually believe testimony is enough. We’re selectively believing what fits our tradition and skeptical about the rest.

If you apply skepticism consistently across all traditions, Hume’s conclusion stands: testimony has never proven a miracle. Natural laws have never been broken.

What This Means

If theism’s main evidence is miracles, and testimony can’t prove miracles, then testimony can’t prove theism. Theists still have philosophical arguments: cosmological, design, ontological arguments. Those are examined elsewhere on this site. But the empirical evidence for theism falls apart under Hume’s logic.

No testimony has been strong enough to overcome the evidence for natural law. None has met the extraordinary standard. Until one does, it’s not rational to believe miracles happened. And without miracles, the case for supernatural beings gets much weaker.

Related Arguments