We were taught that science is a ladder. Scientists observe the world with an open mind, gather data, climb rung by rung and inch their way toward the truth. It is a reassuring, orderly picture – and a false one. Three twentieth-century philosophers (Karl Popper, Thomas Kuhn and Paul Feyerabend) dismantled it, each in his own way, and between them they left a lesson that reaches far beyond the laboratory. They teach us how we know what we think we know, and the humility that should come with any certainty.
Popper was looking for something simple: a clear line between what is science and what is not. He found it in a surprising place. A theory is not scientific because it can be proved, but because it can be refuted. “All swans are white” is a scientific claim because a single black swan sinks it. A theory that explains everything and can never be wrong is not science but something else. For Popper, a good scientist does not try to prove himself right but tests his ideas as hard as he can. We advance by eliminating errors, not by accumulating truths.
Kuhn, a historian, looked at what scientists actually do and found a different picture. Most of the time they do not try to refute anything. They work within a paradigm, a shared framework that nobody questions, solving puzzles and trusting that the theory is sound. When an experiment does not fit, they do not throw out the theory; they assume the mistake is theirs and set the anomaly aside. The anomalies pile up until a crisis comes, and with it a revolution that replaces the whole framework. The shift from a universe with the Earth at its center to one in which the Earth is just another planet is the textbook case: Ptolemy and Copernicus worked within different paradigms. For centuries, astronomers kept adding circles upon circles (the famous epicycles) to Ptolemy’s model to rescue it, until Copernicus came along and, rather than correcting it, replaced it.
Feyerabend took this idea and pushed it to its limit. He combed through the great discoveries in search of the method that had supposedly guided them, and could not find it. His favorite example is Galileo himself, whom legend presents as the hero of evidence. Feyerabend shows that Galileo won by breaking every rule of good method. He argued that the Earth moves when direct observation said the opposite, he trusted a telescope nobody had validated, he relied on hypotheses that had not yet been proved and, above all, he waged a great campaign of propaganda and rhetoric. Had he followed the rulebook, he would have had to concede to his opponents, who at the time had more evidence on their side. Feyerabend’s provocative conclusion, “anything goes”, does not mean that every idea is equally valid, but that there is no rule of method that has not had to be broken at some point for science to move forward.
What unites the three, beyond their differences, is a single, fundamental objection. Scientific knowledge is not an oracle handed down from the heavens, clean and complete, but a human construction – historical and open to dispute – made of bold conjectures, of frameworks that one day collapse, and of decisions in which rhetoric carries as much weight as logic. Science remains the best tool we have for understanding the world. That is precisely why it is worth understanding how it really works, rather than worshipping it as if it were infallible.
In business, none of this is academic. Every organization operates within a paradigm (“that’s how our sector works”, “that’s what our customers want”) that nobody questions because, like Ptolemy’s, it works for a long time. Anomalies appear early, as weak signals that do not fit the official story, and the reflex is always the one Kuhn described: set them aside, put them down to a measurement error, or add another epicycle to the strategic plan. A manager who has taken Popper’s lesson to heart does not look for the data that prove them right, but for the data that could prove them wrong. Anyone who has read Feyerabend knows that the idea that sounds ridiculous in today’s meeting may be the next paradigm shift, and that stifling dissent is the surest way to miss the coming crisis.
For learning, the implications go even deeper. Studying is not memorizing a catalog of settled truths, but understanding how they were built, why they were accepted and what could bring them down. A student who knows that science advances by overturning its own frameworks does not learn less science but more – and, above all, learns to think. In a world where information is limitless and cheap, this is the skill that really sets people apart. It is not about accumulating answers, but about knowing how to weigh them, put them in context and test them.
We should not idealize this, though. It is easy to slide from “no method is sacred” to “everything is opinion and nothing is true”, and here Feyerabend leaves the door wide open to anyone who wants to deny what is well established. Kuhn himself spent half his life explaining that he was not a relativist. The lesson of all three is not that science is worthless, but quite the opposite: it is serious enough, and human enough, to deserve to be understood rather than turned into dogma. Doubting with judgment is not the same as believing in nothing.
This week’s exercise. Think of a “truth” you take for granted in your work or your sector – the kind nobody ever questions. Write it down by hand on a sheet of paper and test it with two questions. First: what evidence, if it came to light, would prove it false? Second: what would be the strongest point in the argument of someone who believed exactly the opposite? The aim is not to change your mind, but to check whether your certainty holds because it is solid or simply because nobody has looked at it closely for a long time.
The great advances have almost never come from people who repeated what everyone else took to be true. In your work or your life, which certainty have you gone too long without really examining?
Sapere aude.

