In the winter of 1610, a man in Padua pointed a homemade tube of glass at Jupiter and saw four little stars he couldn't explain. Galileo Galilei watched them for night after night, and slowly the impossible thing dawned on him. The little stars moved. Not randomly — they swung from one side of Jupiter to the other and back, like beads sliding on an invisible wire. They weren't stars at all. They were moons, circling another planet, four worlds he could see plainly and would never, ever touch.
Here's the puzzle that sits under that whole scene, and it's bigger than Galileo. How can you be certain of something you can never reach, never hold, never poke with a stick? That question is the engine of all of science, and it's what this chapter is built around. Because the honest answer is that science isn't a pile of facts somebody confirmed by touching them. It's a method for being confident about things you can't touch — and that method has a structure worth understanding.
Start with the loop at the center of it all. You observe something. You make a guess about why it's happening. From that guess, you predict something you haven't seen yet. Then you go check whether the prediction comes true. Observe, guess, predict, check — and then around again, because the check almost always sends you back to revise the guess. Galileo did exactly this. He observed the moving dots. He guessed they were moons in orbit. He predicted where they'd be the next night. And the next night, there they were, right where the guess said they'd be.
Notice the part that's easy to skip. The prediction has to be the kind of thing that could come out wrong. If Galileo had said "the dots will do something," that's not a prediction — it's unfalsifiable mush. His guess was strong precisely because it stuck its neck out. It said: tomorrow, this specific dot will be on this specific side. That's the whole game. A claim that can't be tested, that can't possibly fail, isn't worth much in science — not because it's necessarily false, but because there's no way to find out either way. It just sits there, immune to evidence.
Now here's a distinction that trips up almost everyone, and getting it backwards is one of the most common mistakes people make about science. The difference between a theory and a law. In ordinary speech, "theory" means a hunch — "I have a theory about why he's always late." So when someone says evolution or gravity is "just a theory," they think they're saying it's a guess. They've got it exactly upside down.
A law, in science, describes what happens. It's the pattern, often written as a tidy relationship. Drop something and it accelerates toward the ground at a steady rate — that's describable as a law. It tells you the what, reliably, every time. But a law doesn't tell you why. It doesn't explain the underlying mechanism. That's the theory's job. A theory is the big, tested explanatory framework that says why the pattern exists in the first place. As the OpenStax College Physics text puts it, science consists of the theories and laws that are the general truths of nature — two different tools, side by side. Neither outranks the other. A theory doesn't "graduate" into a law after enough proof. They're answering different questions. The law says what; the theory says why.
So when you hear "it's just a theory," the accurate translation is closer to: "it's just a comprehensive, tested explanation that's survived decades of attempts to break it." Which is the opposite of a weak claim. That's the part worth carrying out of this whole section — in science, a theory is the strongest thing you can say, not the weakest.
Now, stay with this for one more step, because there's a third tool in the kit, and it's the strangest of the three. The model. A scientific model is a deliberate simplification — and here's the surprising part — a good model lies on purpose. It's wrong about some things so it can be right about the thing that matters.
Think about a subway map. The famous London Underground map, or the New York City subway diagram. It's geographically a lie. The distances are wrong, the angles are wrong, the rivers are squished, the stations aren't really evenly spaced. If you tried to use it to walk above ground, you'd get hopelessly lost. And yet it's one of the most useful diagrams ever made — because it answers the only question a rider actually has. Which line do I take, and where do I change trains? It throws away geography to nail the journey. That's exactly what a scientific model does. It strips away the messy detail that doesn't matter for the question you're asking, so the part that does matter snaps into focus.
This is where most people get uneasy, because it sounds like cheating. If the model's wrong, why trust it? But that's the whole point — every model is wrong about something, and the skill of science is choosing which wrongness you can afford. A model of a swinging pendulum that ignores air resistance is "wrong." It's also accurate enough to land a spacecraft. The question is never "is this model true?" The question is "is this model useful for what I'm trying to do?" Keep that in your pocket, because later in this course you'll meet a model of the atom that's mostly empty space, and a model of gravity as falling that never quite hits the ground — both wrong in the subway-map way, both exactly right for the journey.
So far this has all been about the tools — loop, law, theory, model. But there's a deeper question underneath, and it's where two famous thinkers go to war. What actually makes something science in the first place, versus astrology, or just storytelling?
The cleanest answer came from Karl Popper, the philosopher who watched the science of the early twentieth century and asked what separated the real thing from the fakes. His answer was falsifiability. A claim is scientific only if it sticks its neck out far enough to be proven wrong. Not proven right — proven wrong. Popper's point was sharp and a little brutal. A theory that explains everything, that bends to fit any possible observation, explains nothing. It can't be tested, because nothing could ever contradict it. Real science, for Popper, makes risky predictions. It says "this will happen, and if it doesn't, I'm cooked." Einstein's theory of gravity predicted starlight would bend by a precise amount during an eclipse. If the light hadn't bent, the theory was dead. That willingness to die is what makes it science. Galileo's moons, again — the guess that could have been wrong tomorrow night is the guess that earned the right to be believed.
Here's a gut-check before we go on. If a friend told you they had a theory that perfectly explained every possible outcome of tomorrow's weather — sun, rain, snow, all of it, no matter what happens — would you call that a strong theory? … By Popper's standard it's the weakest possible one. A theory that can't be wrong tells you nothing, because it never risked anything.
But Popper described how science should work in the ideal. A historian named Thomas Kuhn looked at how it actually works, and found something messier. Kuhn's 1962 book, The Structure of Scientific Revolutions, argued that science doesn't advance by scientists cheerfully discarding theories the moment a prediction fails. In real life, scientists work inside what Kuhn called a paradigm — a big shared framework everyone takes for granted, like the rules of a game. Most of the time they're doing what Kuhn called normal science: solving puzzles inside the paradigm, not testing the paradigm itself. And when an experiment comes out wrong, the first move usually isn't to throw out the whole framework. It's to assume the experiment was botched, or to patch the theory. People defend the paradigm, sometimes for years.
Then the anomalies pile up. The patches stop working. Tension builds, until the whole thing cracks at once and a new paradigm takes over — what Kuhn called a scientific revolution, a paradigm shift. Galileo's moons were exactly this kind of crack. They didn't fit the old picture where everything circles the Earth. They helped tip an entire worldview. Kuhn's science doesn't glide forward; it lurches.
So which one's right — Popper or Kuhn? This is the contested debate at the heart of how science understands itself, and serious people still line up on both sides. Popper's defenders say falsifiability is the only honest standard, the thing that keeps science from collapsing into fashion and consensus. Kuhn's defenders point out that Popper's tidy story just doesn't match the historical record — scientists demonstrably don't abandon theories on the first failed test, and they're often right not to. The most useful reading, and the one this course leans toward, is that they're both true at different altitudes. Popper describes the logic — what a claim has to be to count as science at all. Kuhn describes the sociology — how communities of actual humans, with careers and egos and shared assumptions, actually change their collective mind. One's a rule about evidence. The other's a story about people. They're not really competing; they're zoomed to different levels — which, you may notice, is exactly the move this whole course is built on.
Let's gather the thread before the last turn. Science is a loop — observe, guess, predict, check. A law tells you what happens; a theory tells you why, and "just a theory" gets that exactly backwards. A model is a deliberate, useful lie, wrong about geography to be right about the journey. Popper says real science has to be falsifiable. Kuhn says real science lurches forward in revolutions. And both can be true at once.
Which leaves the biggest bet of all — the one that makes everything in this course worth your time. The OpenStax physicists put it plainly: in the face of overwhelming detail, science has found that a surprisingly small and unified set of laws can explain what we observe. That's not a guarantee. It's a wager nature keeps paying out. A bag of chips and a car battery look like they have nothing to do with each other, but one rule — energy changes form but is never lost — ties them together, and ties them to food calories and heat and light and the spring in a watch.
So the question this chapter has been quietly circling is the question that runs through the rest of the course. Not "what are the facts?" but "what's the smallest set of rules that makes all the facts fall into place?" Galileo couldn't touch his moons. He didn't need to. He had a method for being sure — and that method is about to start pointing at things much closer to home than Jupiter, starting with something you do every time you carry a full cup across a room without spilling it.