Why do extraordinary claims need extraordinary evidence?

Where did the idea come from?

The idea is centuries older than the famous phrase. In his 1748 essay “Of Miracles,” David Hume argued that we should weigh a report against all of our past experience. His rule was that no report of a miracle should convince us unless it would be even more surprising for the report to be false. The Stanford Encyclopedia of Philosophy notes that this maxim is still open to debate about what exactly it means.

In the early 1800s, the mathematician Pierre-Simon Laplace put a similar thought in terms of probability. He wrote that the more extraordinary an event is, the stronger the proofs it needs. His example was an urn holding a million balls, all white except one black ball.

The modern wording came later. Quote Investigator traces “extraordinary claims require extraordinary proof” to a 1975 letter by the sociologist Marcello Truzzi (a citation the site says it has seen only in a snippet and not yet fully verified). Carl Sagan was quoted using the “evidence” version in The Washington Post in 1977, in a story about UFO claims, and printed it in his 1979 book Broca’s Brain. Quote Investigator also found similar sayings going back to 1708, so the idea had been around long before Sagan gave it its famous form.

How does it work, in plain words?

Two things decide how much we should believe a claim. The first is how likely it was before the new evidence came in. Scientists call this the prior, or the base rate. The second is how strong the new evidence is. Strong evidence is evidence that would be hard to explain if the claim were false.

A branch of math called Bayesian reasoning combines the two. You start with the odds you had before, then raise or lower them by the strength of the new evidence. You don’t need the equation to use the idea. A claim that fits everything we know needs only a little evidence. A claim that clashes with a century of careful tests needs a lot.

This also shows why the rule is not a wall. The Stanford Encyclopedia describes a reply to Hume by the mathematician Charles Babbage. If witnesses are more often right than wrong, and truly independent of each other, enough of them can outweigh even a very unlikely claim. The catch is the word “independent.” Ten people repeating the same viral video count as one source, not ten.

What does history show?

Claims that won

Stones from the sky. Before 1803, reports of falling stones were usually doubted by scientists, according to Arizona State University’s Buseck Center for Meteorite Studies. Then, on April 26, 1803, more than 3,000 stones fell near L’Aigle, France. The young scientist Jean-Baptiste Biot went to investigate. He compared the stones with local rocks, checked them against earlier meteorites, and questioned many kinds of witnesses. After his report, the existence of meteorites was accepted.

Moving continents. In 1912, the German meteorologist Alfred Wegener proposed that the continents were once joined and had drifted apart. The U.S. Geological Survey explains that his idea was not well received, partly because he could not explain what force moved them. After his death, new evidence from exploring the ocean floor, starting in the 1950s, revived his idea and led to the theory of plate tectonics.

A germ behind ulcers. In 1982, when Barry Marshall and Robin Warren discovered the bacterium Helicobacter pylori in the stomach, stress and lifestyle were seen as the main causes of ulcers. In his Nobel autobiography, Marshall recalls that his results were disbelieved because they “simply could not be true.” He even infected himself to make the case. Treatment studies by Marshall, Warren and others then showed that killing the bacteria cured the ulcers, and the two men shared the 2005 Nobel Prize.

Claims that failed

Cold fusion. In March 1989, Stanley Pons and Martin Fleischmann at the University of Utah claimed that fusion had happened in a lab experiment that passed electricity through heavy water. Many labs around the world tried to repeat it. A U.S. Department of Energy panel reported that November that most groups found negative results, and that labs claiming extra heat did not find the fusion products that should have come with it. It called the evidence for a new nuclear process “not persuasive,” while noting that some observations were not yet ruled out.

Faster-than-light neutrinos. In September 2011, the OPERA experiment reported neutrinos arriving at Italy’s Gran Sasso lab, 730 km from CERN, a little faster than light. The team asked other scientists to check its work. In June 2012, CERN reported that four experiments measured speeds consistent with light. The result was traced to a faulty part in a fiber-optic timing system. Our story Can Anything Travel Faster Than Light? has more.

What is the strongest objection?

The strongest objection is that the rule can be used to protect old ideas. In a 2016 paper in the journal Philosophia, David Deming of the University of Oklahoma points out that Sagan never defined “extraordinary.” That gap lets people call any idea extraordinary simply because most experts disagree with it. Deming argues that a claim is extraordinary only when a large body of closely matching experience stands against it. A new idea is not extraordinary just because it is new.

This is a fair warning, and Marshall’s story shows why. But notice what finally settled each case above: repeated tests by independent teams. The rule works best as a question (what would convince us?) rather than as a verdict.

What do people often get wrong?

“You can’t disprove it, so it might be true.” Almost anything might be true. The question is whether there is good reason to think it is. Many claims can never be fully ruled out, so “nobody has disproven it” gives no reason to prefer one of them over another.

“The skeptic has to prove me wrong.” The burden of proof falls on the person making the new claim. In Broca’s Brain, Sagan wrote that it should fall on those who make such proposals. That is not an insult. It is where the missing evidence lives.

“Skeptics just say no.” Healthy skepticism means asking for evidence and changing your mind when it arrives. Cynicism means refusing to believe anything, whatever the evidence. Sagan himself, as Deming notes, warned that scientists who shut out new ideas are as much a problem as people who believe too easily. For UFO reports, our story on what official UAP reports say shows the balance in action.

Check it yourself

You can use four steps on any surprising claim, from a video to a headline:

1. What was observed? Separate the raw facts (a light, a reading, a result) from the story told about them.
2. What was inferred? Ask what the claim says the facts mean, and what else could explain them. OPERA’s real cause was a faulty part, not new physics.
3. Does the evidence support it? Look for independent repeats, different methods, and experts outside the original team.
4. What is unresolved? Say what we still don’t know. “Not explained yet” is an honest answer.

Our guide How to Check a Viral Science Claim in Five Minutes turns these steps into a quick routine.