The Complete Guide to How Science Actually Works

1. When Being Old Made You Right

Two thousand years in which the way to settle an argument was to find the right book.

Here is a question that sounds easy: how would you prove that heavy things fall faster than light ones? For nearly two thousand years, educated Europeans knew the answer. They did not go outside and drop anything. They looked it up — because Aristotle had said so, and Aristotle was the cleverest person who had ever lived, and checking a genuinely clever person seemed like arrogance.

It is easy to sneer at this, and wrong. Aristotle was an outstanding observer whose biology was extraordinary. His physics matched everyday experience: push something, it moves; stop pushing, it stops. A feather really does fall more slowly than a stone. If you have never seen a vacuum, air resistance is invisible, and the wrong theory fits the world you can see.

The universe those scholars inherited was Aristotle's improved by the astronomer Ptolemy: the Earth motionless at the centre, the Sun, Moon and planets carried around it on nested spheres, and the whole thing enclosed by fixed stars. It explained the sky, it matched common sense — the ground is not moving — and by the Middle Ages it had been carefully woven into Christian theology.

Geocentric model: The Earth-centred universe. To make it match the observed wobbles of the planets, astronomers added epicycles: small circles carried on the big circles. It worked well enough to predict planetary positions for over a thousand years, which is why it survived.

The geocentric system was not stupid. It made testable predictions and got them roughly right. Its real problem was that keeping it accurate required adding more and more epicycles — a theory that needs a new adjustment every time you look harder is telling you something.

What just happened

• Aristotle's physics and Ptolemy's astronomy dominated European thought for centuries
• Both matched everyday observation and had been integrated with theology
• Arguments were settled by consulting authorities rather than by experiment

You cannot understand how radical early scientists were unless you see that the thing they were attacking was reasonable, useful and respectable.

Then a cautious Polish churchman worked out that moving the Sun to the middle made the mathematics far simpler — and sat on the result for thirty years.

2. Moving the Earth

Copernicus, Kepler, and a claim that felt absurd.

Nicolaus Copernicus finished his heliocentric model — Sun at the centre, Earth as one planet among others — decades before he allowed it to be printed in 1543, the year he died. His motive was not rebellion but elegance: putting the Sun in the middle explained why planets sometimes appear to travel backwards across the sky without inventing extra machinery.

The objections were not stupid either, and this is the part usually skipped. If the Earth is hurtling round the Sun, why do we not feel it? Why does a dropped stone land at our feet rather than behind us? And the strongest one: if we move, the angle to nearby stars should change over the year — an effect called parallax — and nobody could detect any. The answer, that the stars are unimaginably far away, sounded like special pleading. It was true, and parallax was not measured until 1838.

Johannes Kepler (1571–1630)

Spent years trying to force the data into circles, then had the courage to abandon them.

Kepler inherited decades of superbly accurate naked-eye observations from the astronomer Tycho Brahe. Mars would not fit a circular orbit — it was off by eight minutes of arc, an error smaller than the width of a fingernail at arm's length. Kepler trusted the data over the tradition, and discovered that orbits are ellipses. Everyone from Aristotle onwards had assumed circles were perfect and therefore correct.

That is the moment the modern habit appears. Eight minutes of arc was well within what any earlier astronomer would have shrugged off. Kepler treated a small, stubborn discrepancy as more authoritative than two thousand years of assumption — and it broke the whole system open.

Commonly believed: Everyone before Columbus thought the Earth was flat.

Actually: No educated person in Europe had thought that for well over a thousand years; the Greeks had measured the Earth's circumference with reasonable accuracy. The nineteenth century invented the flat-Earth myth largely to make a story about science defeating religion. Real history is messier and more interesting than the cartoon.

What just happened

• Copernicus published a Sun-centred model in 1543, mainly for its simplicity
• Strong observational objections — no felt motion, no measurable parallax — kept it disputed
• Kepler followed Tycho's precise data to elliptical orbits, abandoning perfect circles

Heliocentrism won by degrees and against genuine evidence-based objections. That is what real scientific change looks like: not a revelation, but an argument fought over decades.

Meanwhile in Italy, a professor pointed a new instrument at the sky and saw something the old system said could not exist.

3. The Man Who Looked

Galileo, the telescope, and evidence that could be checked by anyone.

Galileo Galilei (1564–1642)

Did not invent the telescope, but was the first to publish what he saw through one.

Galileo heard of a Dutch device that made distant things look nearer, built a far better one within months, and turned it upwards. He was also a superb self-publicist who wrote in Italian rather than Latin, so that merchants and courtiers could read him. That decision made him famous, and it made his enemies.

What he saw, and why each observation mattered

• Mountains and craters on the Moon — the heavens were supposed to be perfect and unchanging; this was a rocky, damaged landscape
• Four moons orbiting Jupiter — proof that not everything goes round the Earth
• The phases of Venus — full, gibbous, crescent, exactly as a Sun-centred model predicts and the Earth-centred one cannot produce
• Countless unseen stars in the Milky Way — the universe was far bigger than anyone had assumed

The phases of Venus are the decisive item. Ptolemy's arrangement makes certain phases impossible. Galileo did not merely offer a better story — he produced an observation the old model could not accommodate at all, and anyone with a telescope could check it. Verifiable public evidence is the whole of the scientific revolution in one line.

He also demolished Aristotle's physics with experiments rather than argument. Rolling balls down inclined planes to slow the motion enough to time it, he showed that objects accelerate uniformly regardless of their weight, and that a moving body keeps moving unless something stops it. Friction, not nature, is what makes things halt.

The famous story of dropping two balls from the Leaning Tower of Pisa was written down by a devoted biographer and appears nowhere in Galileo's own accounts. His actual method — inclined planes and a water clock — was cleverer, because it made a fast event slow enough to measure with the instruments he had.

What just happened

• Galileo built an improved telescope and published his observations in Italian
• Jupiter's moons and the phases of Venus contradicted the Earth-centred model
• Inclined-plane experiments overturned Aristotle's account of falling and motion

He changed what counted as an argument. Not 'the authorities say' but 'here is what I measured, and you can go and measure it too'.

In 1633 he was summoned to Rome. What happened there is the most misunderstood episode in the history of science.

4. The Trial

What Galileo was actually condemned for.

The Church was not uniformly hostile to Copernicus. Jesuit astronomers were among the best in Europe, several cardinals were interested, and the future Pope Urban VIII had written a poem in Galileo's honour. In 1616 the Church's position was that heliocentrism could be discussed as a mathematical hypothesis but not asserted as physical truth without proof — and, awkwardly, Galileo did not yet have proof. His best argument, that the tides were caused by the Earth's motion, was wrong.

In 1632 he published a Dialogue Concerning the Two Chief World Systems: a conversation between three characters, one defending the old view. He gave that character the Pope's own favourite argument — and named him Simplicio, which in Italian reads uncomfortably close to 'simpleton'. Urban VIII, then under intense political pressure during the Thirty Years' War, took it as a public humiliation.

22 June 1633 — The convent of Santa Maria sopra Minerva, Rome

Galileo, sixty-nine and ill, kneels and reads a prepared abjuration renouncing the view that the Earth moves. He is sentenced to imprisonment, immediately commuted to house arrest, where he spends the remaining nine years of his life. The line 'and yet it moves' is not recorded at the trial and was almost certainly attached to the story a century later.

Commonly believed: The trial was simply religion against science.

Actually: It was that, and also a quarrel about who had the authority to interpret scripture, a personal insult to a proud pope, and a political crisis in which Urban could not afford to look weak. Historians of science stress this not to excuse the verdict — it was a disgrace, and the Church took until 1992 to formally close the matter — but because 'religion versus science' is too small a box for what happened.

The scientific effect was real. Italian science, which had led Europe, went quiet. The centre of gravity moved north to the Netherlands, France and England, where a young Isaac Newton would be born the year Galileo died. And a lesson was learned that outlived the case: claims need to be defensible on evidence, and a scientific community needs to be able to argue without one authority ending the discussion.

What just happened

• The 1616 ruling allowed heliocentrism as hypothesis but not as asserted fact
• Galileo's 1632 Dialogue put the Pope's argument in the mouth of 'Simplicio'
• He abjured in 1633 and lived under house arrest until his death

The trial shows what happens when a single authority can end an argument — and it is the reason the scientific community that grew up afterwards was built on published, checkable, contestable claims.

So what does separate a scientific claim from a merely confident one? The best answer took another three hundred years to formulate.

5. What Actually Makes It Science

Falsifiability, replication, and why being unbeatable is a bad sign.

You were probably taught a method: observe, hypothesise, experiment, conclude. It is a useful summary and a poor description. Real research loops, doubles back, follows accidents, and often starts from a hunch nobody can justify. What holds it together is not a sequence of steps but a set of standards for what counts as knowing something.

Falsifiability: The philosopher Karl Popper's proposal: a claim is scientific if you can say, in advance, what observation would prove it wrong. 'All swans are white' is scientific — one black swan kills it. A theory that can absorb any possible result and explain it afterwards is not strong. It is unfalsifiable, which is a weakness dressed as a strength.

The other pillar is that science is a social process. A result that exists only in your notebook is not yet knowledge. It becomes knowledge by being published in enough detail for strangers to repeat it, reviewed by people who would enjoy finding your mistake, and then either surviving or not. The individual scientist is fallible and biased; the system is designed to be less so than its members.

The machinery, and what each part is for

• Peer review — experts try to find the flaw before publication. Slow, imperfect, and better than nothing
• Replication — someone else repeats it. If it only works in your lab, it is not yet a fact about the world
• Controls — a comparison group, so you can tell whether your treatment did anything at all
• Blinding — neither participant nor experimenter knows who got what, because expectation genuinely changes what people report and what researchers record
• Publishing your methods — the claim has to be checkable, or it is just an assertion with a graph

Since around 2011, psychology, medicine and several other fields have gone through a 'replication crisis': large projects re-ran famous experiments and found that a substantial share did not hold up. It sounds like a scandal, and it is — but notice which discipline discovered it. Science found the problem by running its own checking machinery on itself.

This is also why the word theory causes so much confusion. In ordinary speech it means a guess. In science it means an explanatory framework that has survived enormous testing and unifies many separate observations — evolution, plate tectonics, general relativity. A theory is not a hypothesis that has been promoted; it is a different kind of thing entirely.

What just happened

• The 'scientific method' taught in schools is a simplification of messier real practice
• Popper proposed falsifiability as the test of a scientific claim
• Peer review, replication, controls and blinding are the machinery that makes it collective

Science is reliable not because scientists are unusually honest, but because the system assumes they are not and checks accordingly.

One question remains, and it is the one that decides how you read a headline: how do you tell good science from something that is merely wearing its clothes?

6. Using This Tomorrow

How to read a claim about the world without a laboratory.

You will meet scientific claims constantly and you will almost never be in a position to test them. That is fine — nobody tests most of what they know. What you can do is judge the machinery behind a claim, which is a skill, and it is teachable.

Take an ordinary example. A headline says that a particular food reduces your risk of a disease by fifty per cent. That sounds enormous, and it might be almost nothing. If the risk was two in a thousand and is now one in a thousand, the halving is real and the difference to you is one person in a thousand. The trick is called relative risk, and it is not usually dishonest — it is simply the number that makes the better headline. The question that defuses it is always the same: fifty per cent of what?

Or take the study that finds people who drink red wine live longer. Perhaps wine helps. Or perhaps people who can afford wine regularly also have better housing, less physically punishing work and quicker access to a doctor — and the wine is a marker of a comfortable life rather than a cause of a long one. Researchers try to strip those factors out statistically, and often say plainly in the paper that they could not fully do so. That sentence rarely survives the journey into the news story.

Questions worth asking about any claim

Does science actually progress towards truth?

• Yes — cumulatively: Predictions get more accurate, technologies built on the theories work, and errors get found and corrected. Aeroplanes fly and vaccines work; that is not a matter of opinion or fashion. Later theories usually contain the earlier ones as special cases.
• It shifts rather than accumulates: Thomas Kuhn argued that science works within a paradigm until anomalies pile up, then flips to a new one that is not simply the old one plus more. Newton was not slightly wrong about gravity — relativity describes a different kind of universe. Progress is real, but 'closer to the truth' is harder to defend than it looks.

The reason this matters beyond exams is simple. Every serious decision a society makes about health, energy, food or climate depends on ordinary people being able to tell a well-supported claim from a confident one. That skill — not the list of facts — is what four hundred years of argument actually bought us.

“The first principle is that you must not fool yourself — and you are the easiest person to fool.”

Feynman was describing why controls and blinding exist: not to catch liars, but to catch honest people believing what they hoped to find.

What just happened

• Judging science means judging its machinery, not repeating its conclusions
• Sample size, controls, replication, funding and causation are the standard checks
• Kuhn and Popper disagree about whether science accumulates or shifts

The habits in this chapter are the actual transferable skill. Facts change; the ability to interrogate a claim does not.

Now you know the story

Science is not a list of facts and it is not a five-step method you were taught at eleven. It is a set of habits — check, publish, doubt, let other people try to break it — that took several hundred years and a few ruined lives to establish. Everything in your physics, chemistry and biology courses arrived through that machinery.

• Explain why 'the Earth goes round the Sun' was not obviously true and took a century to win
• Say what makes a claim scientific — and why 'it fits the evidence' is not enough
• Describe what Galileo actually did wrong, according to the people who tried him
• Explain peer review, replication and why science is a social process, not a lone genius one

Back to How Science Actually Works →

The history and philosophy of science is not assessed content on the Edexcel International GCSE science specifications, which assess experimental skills (AO3) rather than the history of scientific method.