So I was having my typically night-time barrage of questions from my 5 year old when the conversation turned to the earthquake drills she has at school. She didn't know what an earthquake was and so I tried to explain plate tectonics to her in an age-appropriate fashion.
After a bunch of good questions like "why aren't there earthquakes all the time if the plates are pushing against each other?" She asked me how we knew that there this pushing was what caused earthquakes if it happened deep in the earth. I said that I don't know how scientists figured this out and she said "can you ask them?" And I promised I would.
So any course 12s out there want to give me an ELI5 [Explain Like I'm 5] for a literal 5 year old on how we know plate tectonics is true? (I could Google it, but I promised her I'd talk to actual scientists).
So, how do we know plate tectonics is true? The first thing is that you can tell that the continents used to be in different places by looking at how their features line up. The eastern coast of Brazil has the right shape to fit right into the Gulf of Guinea, and if you look at all the fossils and the different layers of rock, they way they match up makes sense if a couple hundred million years ago they actually WERE fitted together.
And then you can look along the bottom of the oceans and you find that there are these ridges that run along the middles of the oceans. The seafloor surface is older the further away you get from the ridge, and right at the ridge, there's lava coming to the surface and new rock being formed, and the rocks are brand new.
Elsewhere, you can find these trenches on the ocean floor where it gets really deep, and on the edge of the trenches there are long chains of mountains and islands that are all wrinkly in parallel to the trenches. And you find the trenches on the opposite side of where the ridges are.
So when you put all those pieces together (plus a bunch of other corroborating details, like the rocks on the ocean floor being denser than rocks on land), the explanation that makes sense is that the Earth's surface is broken up into big chunks of solid rock that float on top of the magma inside the earth, and that the chunks move. Seafloor spreading in the Atlantic ocean has pushed the Americas away from Africa, while over in the Pacific, the oceanic crust is getting shoved underneath Asia. (To the northwest of the trenches where that's happening, the Asian plate has wrinkled up like carpet, forming the Japanese island chain.)
Now, an earthquake happens when two chunks of ground slide past one another along a fault in the ground. When the fault hits the surface, you can see that really clearly. There are places in California where you can see a fence that used to be lined up, and then the earth moved, and now on one side of the fault the fence is in one place, and on the other side it's ten feet further that way. Similarly, if one side of the fault slides up and the other one slides down, you can match up layers of different kinds of rock in the ground to see the movement. And there are plenty of people who have watched earthquakes as they happen, so we know that's how they work.
But even if you can't see a fault at the surface, you can figure out where it was and which way the rocks moved by looking at recordings from a bunch of different seismometers, which are instruments that measure ground shaking. It's like how you could tell where a pebble was dropped into a pond by looking at how a bunch of leaves floating on the surface move: the ripples reach the close leaves earlier while they're still big, and they'll reach the leaves that are farther away later on when they're smaller. Seismometers record ripples in the ground, and by tracing the ripples backwards to their origin, you can figure out where the fault was.
If you look at seismometer readings from all over the world, you see patterns that match up with the theory of plate tectonics. First thing, there ARE earthquakes happening all the time. It's just that most of them are so small that we don't notice them without a sensitive scientific instrument. Second, if you draw them all on a map, they outline the edges of the tectonic plates. And third, if you look at the orientation of the earthquake faults, they're consistent with the plates moving as a unit. The Pacific plate is moving sort of to the northwest, away from the mid-ocean ridge in the southeast Pacific, so the earthquake faults in California are all oriented side-to-side because there it's sliding past the west coast of the North American plate, but they're all oriented in a dipping down and under way up along the northeast coast of Asia and along the Aleutians.
So in fact, it's not that we know that the pushing is happening and that's what causes earthquakes, it's actually the other way around: we know that earthquakes happen when two pieces of rock move past one another, and the picture formed by all the earthquakes taken together tells us that there must be pushing going on deep inside the Earth to make them all line up the way they do.
(Wikipedia has lots of good illustrative pictures of all this.)
Don't ever say I never taught ya nothin'.