Conjectures and Refutations

by Karl Popper · 8 ideas

  1. Conjectures and Refutations

    Knowledge grows by guessing and then trying to prove the guess wrong. Theories that survive hard tests are kept for now, never proven forever. Treat your ideas as guesses to test, not truths to defend.

  2. Falsifiability

    A claim is scientific only if some possible observation could show it is false. If nothing could ever count against it, it says nothing testable about the world. Ask of any claim: what would prove this wrong?

  3. Cheap Confirmations

    You can find supporting evidence for almost any idea if you look for it. Support only counts when it comes from a risky prediction that could have failed. Look for tests your idea could fail, not for examples that agree with it.

  4. Explaining Everything

    A theory that fits every possible outcome seems strong but explains nothing. If it can account for one behavior and its opposite, it forbids nothing. Be suspicious of ideas that are never wrong.

  5. Ad Hoc Rescue

    When a theory fails a test, people often patch it with an excuse so it survives. Each patch makes the theory harder to test and less useful. Accept a failed test rather than inventing reasons it does not count.

  6. No Ultimate Source

    Asking where an idea came from, or who said it, does not tell you if it is true. Every source can be wrong. Judge an idea by how well it stands up to criticism, not by its origin.

  7. Observation Needs a Question

    There is no neutral way to just observe. You always look with some question or expectation in mind, which decides what you notice. State your question first, so you know what you are looking for.

  8. Critical Tradition

    Traditions are useful starting points, not things to follow blindly or throw away. The rational move is to keep them while openly criticizing and improving them. Respect inherited ideas and still question them.

More on thinking