“Nature is relentless and unchangeable, and it is indifferent as to whether its hidden reasons and actions are understandable to man or not.” -Galileo Galilei
All of science is rooted in the idea that natural phenomena can be explained naturally, and that if we want to know how anything in the Universe works, all we need to do is ask the Universe the right questions, and the answers will appear.
So what about the question of the night sky, and why it appears to rotate the way it does?
There are two straightforward explanations for this, and from observing the apparent motion of the night sky alone, they’re indistinguishable from one another.
- The entire sky — and all the stars in it — spins around the Earth with a period of 24 hours, causing the stars to change position as we observe it from Earth.
- The entire sky is — to the best of our observations — stationary, and appears to spin because the Earth is rotating beneath it.
These two scenarios, although they would both adequately explain this phenomenon, are vastly different from one another.
But the stars appearing to rotate about the celestial pole is not the only observation we have. By making other observations and interpreting them in the context of these two very different models, we can help determine whether one is superior to the other.
The first clue to which one is correct came back in 1610, when Galileo discovered that the planet Jupiter had satellites of its own orbiting it, and today happens to be the 404th anniversary of the discovery of the fourth (and final) Galilean satellite of Jupiter: Ganymede, the Solar System’s largest moon.
But although that’s suggestive in the sense that it tells us that there are objects orbiting bodies other than Earth, it doesn’t tell us whether the Earth is rotating or stationary.
So what can we see that can give us a clue to whether it’s the Earth or the sky that’s moving? Although the stars always appear to make this rotational motion throughout the night, the stars that are visible in the night sky — as well as their locations — vary greatly throughout the year.
Given that we’ve got the entire sky to consider, this has to do with the position of the Sun. When the Sun appears during the day in the summer, the winter constellations are obscured by the sunlight bathing our atmosphere, and when night falls, the summer constellations are visible. Conversely, the summer constellations are obscured by the Sun during days in Winter, while the winter constellations are visible at night.
Again, both models can accommodate this, but they look very different.
If the Earth is truly stationary, then the Sun would have to move to different locations relative to the night sky throughout the year. In addition to its once-a-day orbit around the Earth, it would have to migrate in one additional circle relative to the background stars each year, in order to explain why the visible constellations vary throughout the seasons.
On the other hand, if the Earth is allowed to move, then it can also move around the Sun, explaining why different constellations appear in the night sky at different times of the year.
We also need to explain the annual changes that occur in the Sun’s path.
From the point-of-view of us here on Earth, particularly those of us who live well away from equatorial latitudes (outside of the tropics), the Sun’s path through the sky varies significantly throughout the year.
The lowest the Sun ever appears — at zenith — above the horizon occurs during the winter solstice, while its highest point occurs during the summer solstice.
In the Earth-is-stationary model, the Sun needs to change its location in the sky significantly throughout the year: in addition to its once-a-day journey around the Earth, it needs to change its location relative to the celestial sphere by a whopping 47 degrees every six months. Why the Sun moves in this path so slowly relative to the celestial sphere but so quickly relative to Earth is not explained by any predictive mechanism in this model.
On the other hand, if the Earth is allowed to move, this would result simply from the Earth moving around the Sun while it rotates on its tilted axis. If the Earth is the moving thing, its rotation and its revolution are allowed to be separate quantities, which could explain the vastly different timescales for days (the period of Earth’s rotation) and years (the period of Earth’s revolution).
In other words, we can invent mathematics to explain these observations in the “Earth is stationary model,” but there’s no physical mechanism — no governing theory — to explain why or how this happens. In the “Earth is rotating” model, however, so long as we accept that revolution and rotation can be separate properties of objects, this is easy to explain. Again, both models are still allowed, but the complexity and power of each explanation is different. Let’s throw in just one more object: the Moon.
Much like the Sun, the Moon follows a very similar path throughout the sky: it rises towards the East, sets in the West, and it does this — rises and sets — once per day. It also appears to migrate relative to the stars, completing an extra circle about once every 29-to-30 days.
The big difference between the Moon and the Sun is noticeable when there’s a Full Moon.
While the Moon never varies by more than 5 degrees from the Sun in terms of its inclination to Earth, there’s a huge seasonal difference between the Full Moon and the Sun. When the Sun reaches its maximum height above the horizon, during the summer solstice, the Full Moon achieves its minimum height above the horizon. And when the Sun is at its minimum height during the winter solstice, the Full Moon reaches its maximum height above the horizon!
If the Earth must remain completely stationary, we must again put the Moon’s orbit in, making an extra circle relative to the celestial sphere every lunar month, and inclined at nearly the same (but not quite) the same amount relative to the celestial sphere as the Sun.
We need this, of course, to explain the observed Lunar and Solar eclipses, which can easily be deduced to be to the interplay of shadows between the Sun, Moon and Earth.
But if you allow the Earth to move, you can not only explain the daily motion of the stars, Moon, and Sun relative to the Earth’s sky by the Earth’s rotation, you can explain the lunar and solar motions relative to the rest of the sky as revolutionary orbits due to the force of gravity.
Note that once you allow for gravity — once you discover that there’s a physical law governing the motion of the objects in the heavens — now you can suddenly predict how all these motions occur without having to introduce any new parameters. You give the law of gravity the masses, velocities and positions of each object, and it gives you the only possible solutions for motion: the ones we observe in the Solar System.
If you insist that the Earth remain stationary and the celestial sphere rotates, you can make a working model for the Earth, Sun, Moon and stars, but it requires you to put the motions of the Sun (an extra revolution tilted at 23 degrees relative to the celestial sphere per year) and the Moon (an extra revolution tilted at 5 degrees relative to the Sun per lunar month) by hand, with no physical explanation for these motions.
That was exactly what the Ptolemaic Model did, which adequately described these motions without explaining them. This is why we needed the theory of gravity (and why we need scientific theories in general): to explain why the objects in the sky make the apparent motions that they do, and to explain why the path of the Full Moon nearest the Summer Solstice is — to within that 5 degree tolerance — is identical to the path of the Sun at the Winter Solstice.
The theory of gravitation, and the heliocentric model that goes along with it, is much more predictively powerful, scientifically, than the description that came before it, and the predictions that the theory make are backed up by both terrestrial and celestial experiments and observations.
This is to say nothing of the other observations that have come since, including the phases of Venus,
the latitude-dependent motion of a Foucault Pendulum over the course of a day,
and the advances that have come along with General Relativity, all of which demand a rotating Earth.
But the mere fact that one of the possible explanations makes predictions that have been borne out for what’s going to occur given new planets, moons and orbital systems and the other doesn’t is enough to accept one picture — the one of the moving Earth — over the other. That’s how we can reliably discriminate between scientific theories; in the end, predictive power and accuracy, not beauty or aesthetic, is the ultimate arbiter.
That’s why we not only conclude that the Earth moves, but that’s how it — and the other major objects in our day and night sky — determines what we see in the heavens above.
An earlier version of this post originally appeared on the old Starts With A Bang blog at Scienceblogs.