The Argument from Miracles

The Argument from Miracles says that events which break the laws of nature - healings of terminal illness, resurrections, water turning to wine, the parting of seas - are evidence for a supernatural being who acts in the physical world. Every major religion reports miracles, and if even one is real, then a power that can override natural law must exist. Philosopher Richard Swinburne is the leading modern defender, arguing that miracle testimony can be rational to accept under the right conditions. We score the argument 5/100 for soundness: despite thousands of years of miracle claims, none has ever been verified under controlled scientific conditions, and David Hume’s classic objection still stands.

The Formal Argument

The argument can be structured as follows:

1. There are credible reports of events that cannot be explained by natural causes (miracles).
2. The best explanation for these events is supernatural intervention by God.
3. Therefore, God probably exists.

Both premises are contested. Premise 1 has to show the events really happened and really break natural law - a very high bar. Premise 2 has to rule out every possible natural explanation and show that a supernatural cause is more likely - an even higher bar.

What Counts as a Miracle?

David Hume defined a miracle as a violation of the laws of nature, and that remains the standard definition. Aquinas defined it more broadly as an event beyond the power of any natural agent. Richard Swinburne offered a middle path: a miracle is a one-off exception to natural law, caused by a god.

The definition matters because it sets what counts as evidence. If a “miracle” is just an unusual event we do not yet understand, it does not point to the supernatural - it is just a gap in current knowledge. For the argument to work, miracles have to be truly impossible under natural law, not just rare or unexplained.

Common categories of claimed miracles include:

• Healing miracles: Sudden recovery from terminal disease, regaining sight or mobility, surviving injuries doctors called impossible.
• Resurrection and raising from the dead: The Christian resurrection story, accounts of saints raising the dead, near-death experience reports.
• Nature miracles: Turning water to wine, multiplying food, parting seas, stopping the sun.
• Eucharistic miracles: Bread and wine said to turn into flesh and blood, like the Miracle of Lanciano.
• Marian apparitions: Reported sightings of the Virgin Mary at Lourdes, Fatima, Guadalupe, and other places.

Hume’s Devastating Critique

David Hume’s argument against miracles, in Section X of An Enquiry Concerning Human Understanding (1748), is still the most influential philosophical response. Hume made four main points:

The weight of evidence always favors natural law. All of our experience confirms that natural laws hold without exception. Against this huge body of evidence we have the word of a few witnesses - and witnesses are fallible. It is always more rational to believe a witness was mistaken, deceived, or lying than that a law of nature was actually broken.

Testimony decays as it spreads. Miracle reports are usually second-hand, third-hand, or even further from the event. Even honest witnesses misremember and embellish. The further a story travels from the original event, the less reliable it gets. The most dramatic miracle claims come from ancient texts where checking is impossible.

Human psychology favors miracle belief. People love wonderful, surprising stories. Wonder feels good, and people are wired to believe and pass on amazing tales. That produces a systematic bias toward accepting miracle claims whether they are true or not.

Competing miracles cancel each other. Every religion claims miracles, and the claims rule each other out. The miracles of Christianity, Islam, Hinduism, and Buddhism cannot all be real, because they support contradictory theological conclusions. If Christian miracles prove Christianity true, Islamic miracles prove Islam true - and both cannot be true at once. So miracle claims from different religions count as evidence against each other.

The Scientific Verification Problem

No miracle has ever been verified under controlled scientific conditions. This is the single most important fact in evaluating the argument.

Faith healing. Despite thousands of claimed miracle healings at sites like Lourdes, the evidence does not hold up. The Lourdes Medical Bureau was set up in 1905 to evaluate healing claims. Out of millions of visitors over more than a century, it has recognized only 70 cases as “unexplained” - and “unexplained” is not the same as “miraculous.” No faith healing has ever regrown an amputated limb or produced any result that is clearly impossible under natural law. The healings that are claimed tend to involve conditions (cancer remission, recovery from illness) that have known if rare natural explanations.

Spontaneous remission. Spontaneous remission of cancer happens in roughly 1 in 60,000 to 1 in 100,000 cases. With millions of cancer patients worldwide, hundreds of spontaneous remissions are expected each year by chance alone. Some will happen after prayer, pilgrimage, or anointing - but that is what probability predicts, not evidence of a miracle.

The placebo effect. Placebos produce real changes in the body - less pain, lower blood pressure, immune effects, and even some tumor shrinkage in studies. Rituals, prayer, and the emotional intensity of pilgrimage can act as powerful placebos. When someone is healed after a religious experience, the placebo effect is a plausible natural explanation that has to be ruled out before invoking the supernatural.

Controlled studies of prayer. The largest and most rigorous study of intercessory prayer, the STEP study (Study of the Therapeutic Effects of Intercessory Prayer, 2006), found no significant difference in recovery between patients who were prayed for and those who were not. Patients who knew they were being prayed for actually did slightly worse, perhaps from performance anxiety.

The Documentation Paradox

One of the most telling patterns in miracle claims is that the better the documentation gets, the smaller the miracles become. The most spectacular miracles - seas parting, dead rising, the sun dancing in the sky - come from ancient texts and pre-scientific settings where checking was impossible. As recording technology and science have improved, miracle claims have grown less dramatic, less frequent, and less verifiable.

In the age of smartphones, security cameras, and scientific instruments, no clear, dramatic miracle has been caught on camera or recorded by an instrument. Defenders sometimes say God avoids miracles in places where they could be verified, because that would remove the need for faith. But this defeats the argument: if miracles deliberately dodge verification, they cannot serve as evidence.

This trend fits the natural explanation. What ancient peoples saw as miracles were events they could not explain with the science of their time. As science has grown, the category of “unexplainable” has shrunk to match.

The Catholic Verification Process

The Catholic Church runs the most formal miracle verification process in the world, mainly through the Congregation for the Causes of Saints, which evaluates miracle claims as part of canonization. Two verified miracles are usually required for sainthood.

Medical experts review the documentation to decide if a healing has a natural explanation. The method has serious limits:

• It works by failing to find a natural explanation, not by finding positive evidence of a supernatural one. “We cannot explain it” is not the same as “God did it” - this is the argument from ignorance fallacy.
• The investigations assume miracles are possible, which introduces confirmation bias.
• The bar for “inexplicable” varies, and some approved miracles involve conditions (cancer remission, recovery from coma) with documented natural precedents.
• Independent replication is never attempted, and outside scientists rarely get full access to the medical records.

Swinburne’s Defense

Richard Swinburne, the leading modern defender of the argument, argues that Hume set the bar for miracle evidence too high. Swinburne claims:

• Testimony can be enough for miracles if the witnesses are numerous, independent, trustworthy, and had no motive to lie.
• Background evidence matters: if you already have reason to think God exists (from other arguments), the prior probability of miracles is higher, which makes the testimony more compelling.
• A single well-attested miracle is enough to show supernatural intervention. You do not need repeated miracles under lab conditions.

Swinburne’s approach uses Bayes’s theorem. The chance a miracle report is true depends on the prior probability of God, how reliable the witnesses are, and how likely the event was to happen naturally. If the prior probability of God is not tiny, miracle testimony can rationally shift your beliefs.

Critics say this is circular when used to argue for God’s existence. Swinburne assumes a non-trivial probability of God to argue that miracle evidence is credible, then uses miracle evidence to support God’s existence. The argument only works for people who already give God a significant chance.

Connection to Religious Experience

The Argument from Miracles is closely related to the Argument from Religious Experience. Both treat subjective reports of supernatural encounters as evidence for God. The key difference is that miracles are supposed to be public events - visible to witnesses and in principle detectable by scientific instruments - while religious experiences are private events inside one person’s mind.

This matters because it means miracles should be more testable. A real miracle would leave physical evidence: healed tissue, changed chemistry, documented impossibility. The failure to find such evidence under controlled conditions hurts the Argument from Miracles more than the difficulty of testing private experiences hurts the Argument from Religious Experience.

Our Scoring

Soundness: 5/100. This is among the lowest soundness scores on the site. No miracle has ever been verified under controlled scientific conditions. Hume’s critique - that it is always more rational to doubt testimony than to accept that a law of nature was broken - still holds after nearly three centuries. Spontaneous remission, the placebo effect, misperception, and embellishment give natural explanations for the most credible healing claims. The documentation paradox (miracles shrink as verification tools improve) suggests miracle reports reflect human limits in understanding nature, not actual supernatural intervention. The argument-from-ignorance fallacy - treating “unexplained” as “God did it” - weakens even the most carefully investigated cases.

Personal God: 70/100. The soundness score is very low, but the God probability scores are high - and this is the most important distinction in the scoring. If miracles were real, they would strongly support a Personal God who acts in individual human lives, responds to prayers, heals the sick, and suspends natural law for specific purposes. The very concept of a miracle assumes a personal, caring God who pays attention to people. No other kind of god maps as well onto miracles.

Creator/Designer: 75/100. The Creator score is slightly higher because any being capable of overriding the laws of nature must have either created those laws or have power over them, which is essentially the same as being a creator. A miracle-working being is at minimum an architect of reality who can modify it at will. This fits the creator/designer concept very closely.

Higher Power: 75/100. The Higher Power score matches the Creator score because any supernatural being capable of breaking natural law automatically counts as a “higher power” - a force or consciousness beyond the natural world. All three God probability scores are high because miracles, by their nature, imply something supernatural. The question is entirely about whether miracles actually happen (soundness), not what they would imply if they did. The gap between the 5/100 soundness score and the 70-75/100 God probability scores captures this: the evidence for miracles is extremely weak, but the theological payoff of real miracles would be extremely strong.