Here’s something which puzzled me for a long time, until I happened to come across the answer a couple of years or so ago. I think the book Episodes from the Early History of Mathematics by Asger Aaboe (Cambridge University Press, 1997) provided me with the revelation, but I wouldn’t swear to it.
When people began watching the motions of the Sun, Moon and planets in the sky and measuring them properly—which happened many centuries ago—their movements turned out to be quite complicated, especially if you assumed (as most people did) that the Earth was stationary at the centre of the universe. Everything circled the Earth once a day, but with various wobbles superimposed on the motion.
This led to a rather unsatisfying and complicated picture involving epicycles, in which a planet’s motion in the sky was a combination of two or more movements: its circle around the Earth plus the various wobbles.
Before I go any further, let me point out that the the idea of a flat earth wouldn’t have entered into this: the realisation that the Earth was a sphere happened around 300 BC and Eratosthenes was the first to measure its size—with remarkable accuracy—in about 240 BC.
It was occasionally suggested that things might make more sense if actually everything went round the Sun, with the Earth rotating once a day. Yet, there was great resistance to that explanation. Why? What was the problem? Why not switch to the obvious answer? Why not just do it? Why was the “Copernican revolution” so revolutionary?
One reason was that Copernicus’ system, which based everything on combinations of perfectly circular movements, ended up being even more complex than what had gone before. But there’s another, very simple and logical reason.
If we look at the night sky we see the Moon, and lots of little points of light. Mostly these keep the same positions relative to each other, simply circling the North pole (or South pole if you’re south of the equator). They behave as though they’re attached to a rigid sphere, making them all rotate together. That was called the sphere of the fixed stars, and was thought of as the outermost part of the universe.
A few of them, though, seem not to be attached to it. They move around according to rules of their own, and were given spheres of their own to move them. But apart from that—and from not twinkling—they look pretty much the same as the other points of light. They just happen to be attached to their own spheres instead of the outer one.
How far away were the fixed stars? Nobody knew, but they couldn’t really be much further away than the moving ones or they’d be too dim to see.
So far so good. But why can’t the Sun be at the centre instead of us? What difference would it make, apart from a little bit of wounded pride?
Putting the Sun at the centre of this picture creates a huge, glaring problem.
The closer you are to something, the bigger it looks. If the Earth went round the Sun, then any given constellation would change size in the sky as we moved towards and away from it. This would be clearly visible. Yet we don’t see it. And that, surely, proves that the Earth isn’t moving. It must be the Sun which moves around the Earth.
So if you want the Sun at the centre, you have to explain why the fixed stars always stay the same distance apart in the sky. This entails making the sphere they’re attached to considerably bigger. And then you have to make them bright enough to see, so they’re not little twinkly things any more, but are like other suns in their own right.
But the Sun and its collection of spheres was the entire universe. Saying that the stars are really other suns is like saying that actually there are thousands of universes. It’s mind-boggling. It sounds somewhat insane.
It’s not a matter of demoting the Earth a little from its symbolic position as the centre around which everything revolves, and giving that honour instead to the Sun; it’s a matter of making both the Earth and the Sun utterly insignificant in the scheme of things.
Seen that way, the idea is indeed shocking.