Carl Sagan's baloney detection kit taught us how to separate good science from the work of charlatans. In 2026, that matters more than ever.
  • In his influential book, “Science as a Candle in the Dark,” Carl Sagan wrote about how susceptible humans are to being deceived, and how healthy skepticism can inoculate us against those deceptive tactics.
  • Using lessons from the training that scientists go through, he devised a “baloney detection kit,” which contains 9 lessons that everyone can apply to just about any situation they encounter.
  • Are you being deceived? Is someone attempting to deceive you? How can you tell? And should you pay attention to what they’re saying, or be critical of it? These 9 tools are more relevant than ever.

1.) Demand independent confirmation of whatever statements are asserted as facts.

2.) Encourage substantive debate from all points of view by those with substantial, relevant expertise.

3.) Don’t accept an argument from an authority because that person is an authority. Instead, judge arguments based on the merits of the underlying facts, and how experts scrupulously interpret those facts.

4.) Spin as many hypotheses as you can that are consistent with the data. Every possible explanation that isn’t ruled out or contradicted by the already-existing data should be considered, and each hypothesis should be tested and examined as rigorously as possible.

5.) Whatever your favorite, most preferred hypothesis is — especially if it’s your original idea — be its harshest critic. By attempting to knock it down or poke holes in it as hard as you can, you’ll determine how well it stands up under the steeliest of scrutiny. (And if you don’t, others will.)

6.) Don’t settle for a qualitative analysis of the issue. Be quantitative: ask and answer the key question of “by how much?”

7.) If there’s a chain of argument being put forth, then every link in the chain, from the premise to the final conclusion, must be sound.

8.) The convenient rule of Occam’s Razor: to choose the simplest explanation among multiple hypotheses that explain the data equally well.

9.) Ask whether the hypothesis, at least in principle, can be falsified. Non-falsifiable and untestable hypotheses cannot be checked out, and hence those ideas are incapable of disproof.

Web archive link

  • Lvxferre [he/him]@mander.xyz
    6·
    13 hours ago

    Carl! There’s a dead human in our house! …sorry, wrong Carl.

    Okay, serious now. (The above is a reference to Llamas with Hats.) I think Sagan covered a lot of bases with this, and I heavily recommend everyone here to read The Demon-Haunted World, it’s an amazing book. So I’ll focus on something related: fallacies.

    A lot of people think fallacies are just a “DEBATELORD REEEEEE!!!1” thing. They are not; fallacies are better understood as reasoning flaws. If your reasoning is fallacious, even if you start off with accurate information, sometimes the conclusion will be bullshit. So I think it’s important to identify at least the most common types of fallacy out there; you don’t need to remember the names (they’re just fluff), but if you’re able to smell the fallacy you’ll be way less likely to fall for bullshit.

    Wikipedia has a surprisingly good list of fallacies. In special, I’d recommend people to check the following:

    • begging the question: when the conclusion is assumed to be true, and this is used to build an argument proving the conclusion is true
    • false dichotomy: if bananas aren’t red, then they must be blue~
    • genetic fallacy: invalidating an argument not by its own merits, but based on its origin. It encompasses #3 in Sagan’s list (appeal to authority), but also the opposite (argumentum ad hominem - and no, ad hominem is not a fancy way to convey “insult”, you can be insulting without ad hominem and vice versa).
    • straw man: when the person euthanising a 12yo cancerous dog becomes the “puppy killer”.
    • red herring: might as well call it by its social media names: “ackshyually”, “whataboutism”, etc.
    • ObjectivityIncarnate@lemmy.world
      3·
      12 hours ago

      It’s also important to keep in mind that a fallacious argument leading to a conclusion does not actually disprove the conclusion; identifying the fallacy just means that if the conclusion is correct, that argument is not the path to it. And if the fallacious argument is the only path to conclusion X, then there is simply no basis for presuming X to be correct at all.

      red herring: might as well call it by its social media names: “ackshyually”, “whataboutism”, etc.

      Well, “ackshyually” is actually (:P) rooted in mocking people who pedantically and pointlessly correct others, to the point of being more irritating than informative. The ‘Jimmy Neutron salt’ meme shows a pretty solid example of that kind of behavior:

      Red herrings as a type of fallacy, on the other hand are about using something to support an argument that doesn’t actually have anything to do with it. The Wikipedia page itself gives a solid example of this:

      For example, “I think we should make the academic requirements stricter for students. I recommend you support this because we are in a budget crisis, and we do not want our salaries affected.” The second sentence, though used to support the first sentence, does not address that topic.

      • Lvxferre [he/him]@mander.xyz
        1·
        4 hours ago

        Ah, the “fallacy fallacy”? Got a nice example of that:

        • Alice: “Experts say junk food harms your health. Experts can’t be wrong, so stop eating junk food.”
        • Bob: “This is appeal to authority! It’s fine to eat only junk food.”

        On “ackshyually”. I was trying to be succinct so I got sloppy, and the Wiki does have a good explanation of the fallacy, but basically: if you see an “ackshyually” in social media, it’s usually used for the sake of a red herring.

        That’s because the ackshually isn’t just pedantic and irritating, it’s also distracting. It sounds like someone is contradicting you, without addressing the core argument, you know?