Science isn't a club or a religion. It's a way of checking ideas. You make a guess, test it, see whether the real world agrees, and let other people check your work. When the evidence says you were wrong, you change your mind. That's the whole trick, and it's why science keeps getting better.
Science is a process, not a set of beliefs. It's a toolkit for testing ideas against evidence, built so that mistakes get caught and fixed. A religion asks you to hold certain beliefs; science asks what would show this is wrong? and then goes looking.
"Science is just another religion. Scientists have faith in their theories, the same way we have faith in God."
The cycle of science
There's no single "Scientific Method" that every scientist follows step by step. Real science is messier and more creative than the textbook version.1 But the core loop looks like this:
Fact, theory, belief, faith: sort them yourself
These four words get mixed up constantly. Here's how they're used on this site, following science educators at UC Berkeley:1
| Word | Meaning | Example |
|---|---|---|
| Fact | Something directly observed or measured that anyone can check | Water boils at 100 °C at sea level. |
| Hypothesis | A testable proposed explanation | "This fever is caused by a bacterium." |
| Theory | A broad explanation backed by lots of evidence and many tests | Germ theory; the theory of evolution; the Big Bang theory |
| Belief | Something a person accepts as true, with or without good evidence | "My team will win." "God exists." |
| Faith | Confident belief, especially without evidence that could check it | "Jesus rose from the dead." |
Note: in science, "theory" is a high rank, not a guess. And Berkeley's educators point out that scientists accept ideas based on evidence rather than "believing" them regardless of evidence.1
Tap the label you think fits. Then read why.
Science corrects itself, and that's the point
People sometimes say "science keeps changing its mind, so why trust it?" But changing its mind when evidence changes is exactly what makes science trustworthy. Two real examples from this site:
In 2021 a scientific journal published a paper claiming a space rock exploded over an ancient city near the Dead Sea, possibly inspiring the Sodom story. Other scientists found manipulated images and data problems, and argued the "melted" material could come from ancient metal-working. The journal retracted the paper in April 2025.3 See the Sodom page.
Early radiocarbon dating assumed the amount of carbon-14 in the air never changed. By 1969 scientists knew it does, so they built correction curves using tree rings, corals, and lake-bed layers. The corrections made many dates older, not younger.4 See Dating methods.
Compare that with how many religious traditions handle a contradiction: by finding a reading that makes the text still right. That's the core difference. Science asks "what would prove us wrong?"; dogma asks "how can this still be true?"
The shrinking pocket of "God did it"
For most of history, people explained what they didn't understand with gods or spirits. One by one, those gaps filled with natural explanations. Drag the slider to watch the circle of knowledge grow.
Astrophysicist Neil deGrasse Tyson made this point in an essay called "The Perimeter of Ignorance": "Another practice that isn't science is embracing ignorance," he wrote, criticizing the move from "I don't know how it works" to "so it must be the product of a higher intelligence."2
Many thoughtful believers agree the "god of the gaps" is a bad argument and don't use it. They see God not as a gap-filler but as the reason there are natural laws at all. Berkeley's science educators also note that science studies natural phenomena, so it can't directly support or disprove the supernatural, and that many scientists are religious.1 The atheist response: then the question becomes whether there's any evidence for that kind of God, which is what the rest of this site examines.
The one question that separates them
"What evidence would change my mind?" and has an answer. For example, a rabbit fossil in rocks 500 million years old would wreck the current picture of evolution.
"How can I keep believing this?" and treats doubt as a test to pass rather than information.
Try asking that question about anything you believe, including your doubts. It's the most useful thinking tool there is. More tools: Thinking tools & fallacies.
Sources
- UC Berkeley, Understanding Science, Correcting misconceptions: no single "Scientific Method"; theory, fact, belief vocabulary; science can't support or contradict the supernatural; a 2005 survey found over 48% of scientists had a religious affiliation.
- Neil deGrasse Tyson, "The Perimeter of Ignorance" (American Museum of Natural History), excerpted at The Panda's Thumb.
- Retraction Watch, Sodom comet paper to be retracted (April 2025).
- Science Meets Religion, How reliable is radiocarbon dating?
- The historical examples in the "shrinking pocket" model (lightning, disease, planets, the diversity of life, the age of the Earth) are standard history of science; individual citations still to be added.