Concepts explained
Geocentrism — Fourteen Centuries of Correct Predictions
The sun rises and sets, the stars wheel overhead, and the ground does not stir. Geocentrism was the world as seen, and the calculations built on it worked for fourteen centuries. What brought it down was not a better fit to the data — Copernicus was not more accurate.
Phlogiston was held the wrong way round; the aether was not there at all. This one is different again. Geocentrism worked. Of all wrong theories it was the longest-lived and the most useful.
What ought to be visible if the Earth moved
One misconception first. It is not that the ancients failed to imagine a moving Earth. They imagined it, examined it, and rejected it with reasons.
In the third century BC Aristarchus put the sun at the centre. His proposal was not taken up, and why is the heart of this story.
- If the Earth circles the sun, the angle to a given star must change over six months. However carefully anyone measured, it did not.
- If the Earth were moving that fast, a stone thrown straight up should land behind you. It landed at your feet.
- Clouds and birds are not attached to the ground. Why are they not left behind?
- Heavy things fall towards the centre of the universe. If the Earth is not that centre, why does a stone fall to it?
Mars goes backwards
The real problem geocentrism had to solve lay elsewhere: the planets move strangely.
Mars drifts eastward against the stars, then every couple of years it stops, reverses, and travels west for a while before stopping and resuming. It traces a loop across the sky.
In AD 150 Ptolemy completed the answer in the Almagest: put a small circle on a large one.
Not a fudge but an instrument
It is easy to take epicycles for comic patchwork. They were not.
Stacking circles on circles can imitate any repeating motion as closely as you like. We now know why this works mathematically: anything that repeats can be decomposed into a sum of simple circular motions. Ptolemy did not know the principle, but by hand he was doing the thing.
Two widespread stories
The usual continuation runs: as observation improved, epicycles multiplied into dozens and hundreds until the system collapsed under its own weight.
Not so. The whole Ptolemaic system used about forty circles, and that number did not swell steadily through the Middle Ages. Complexity did not kill this theory.
The second story is more widespread still: that Copernicus abolished epicycles.
What was gained instead
What Copernicus gained was not accuracy but explanation, and the difference lies here.
why do Mercury and Venus never stray far from the sun? A separate rule was needed, tying their epicycle centres to the sun's direction
why does a planet reverse only when it stands opposite the sun? It was simply arranged that way, for no reason
Mercury and Venus run on inner tracks and so cannot stray; retrograde motion is the Earth overtaking an outer planet
A little further in
The data did not decide it
Tycho Brahe pushed naked-eye observation to its limit. His measurements were the finest anyone had.
And he rejected heliocentrism, for the reason given two thousand years earlier: he could not find . If his instruments could not catch it, the stars must be absurdly far away — and that was hard to accept.
So he offered a compromise. The Earth stands still, the sun goes round the Earth, and the other planets go round the sun.
Venus turns full
In 1610 Galileo turned a telescope on the sky. The decisive sight was not Jupiter's moons but the phases of Venus.
Venus waxes and wanes like the moon. But in Ptolemy's system Venus always lies between us and the sun, so it can never show a full face. Galileo saw a full Venus.
Eight minutes of arc
Kepler inherited Tycho's Mars data and tried to fit circles to it. He very nearly succeeded: the discrepancy was eight minutes of arc, about a quarter of the moon's width.
An earlier astronomer would have ignored it. But Tycho's data were not sloppy enough to be wrong by that much.
Because these eight minutes could not be ignored, they alone led to a complete reformation of astronomy.Kepler
In 1687 Newton answered why an ellipse: a single inverse-square force yields all three of Kepler's laws. Heliocentrism was no longer a diagram of the sky but the same physics as the ground.
1838: the last objection answered
The two-thousand-year-old objection still stood, though. Nobody had seen a stellar parallax.
In 1838 Friedrich Bessel caught it at last, in 61 Cygni: three tenths of a second of arc, roughly the angle of a coin a kilometre away.
What it left
decomposing repeating motion into sums of simple circular motions is now a basic tool for signals and waves
a catalogue of 1,022 stars and a method of observing, carried down fourteen centuries as the floor Tycho and Kepler stood on
distrust a theory with many coincidences to explain — an answer to how you choose when accuracy is equal
- 340 BCAristotle — sets out the Earth-centred cosmos
- 270 BCAristarchus — proposes a sun-centred system; it is not accepted
- 150Ptolemy — the Almagest completes the epicycle system
- 1543Copernicus — publishes heliocentrism (keeping the epicycles)
- 1572Tycho — precision observation, and a compromise system
- 1609Kepler — puts in the ellipse and removes the epicycles
- 1610Galileo — the phases of Venus end Ptolemy
- 1687Newton — answers why the orbit is an ellipse
- 1838Bessel — finally measures stellar parallax
The question that remainsCopernicus was right without decisive evidence. Tycho, looking at the same data, was carefully wrong. The difference between them could not be seen at the time and only emerged later. So by what can we tell apart a claim that is right without evidence from one that is merely wrong without evidence?