The colossal movements of tectonic plates shape our world, influencing the composition of Earth’s atmosphere, the planet’s protective magnetic field and perhaps even the flourishing of life. Now researchers have compelling evidence that some form of plate tectonics might have started as early as 3.48 billion years ago, according to a study in Science.
Using magnetic traces from ancient pieces of Earth’s crust, researchers found that a chunk of what is now Western Australia drifted toward the magnetic north pole over a few million years as part of South Africa remained stationary. It’s the earliest documented instance of relative plate motion by more than half a billion years, and it has implications for understanding early life on Earth and how the planet’s tectonic activity began. [Disclosure: The author of this article embedded with the research team during last year’s field season.]
Earth today is a jigsaw of giant chunks of crust that travel across the planet, smashing together like huge bumper cars, pushing up mountain ranges and melting back into magma along their edges. All this activity, called plate tectonics, seems to be unique in our solar system. It’s believed that our rocky neighbor planets instead have a continuous, solid shell.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
No one knows, however, how or when plate tectonics got started on Earth in the first place. “It’s one of the most fundamental questions in Earth science,” says study coauthor Roger Fu, a Harvard university paleomagnetist. Geologists use various tools to investigate the state of Earth’s crust over the eons, but the gold standard is evidence of relative motion: one piece of Earth’s crust moving away from or toward another piece. For that, Earth’s magnetic field—powered by the motion of the planet’s core—holds the key.
This illustrated cross-section of Earth 3.5 billion years ago shows the core generating a magnetic field, as well as subducting tectonic plates.
Alec Brenner, Harvard University/Yale University
Like any magnet, Earth has north and south magnetic poles, aligning roughly with the globe’s geographic poles. These poles flip at irregular intervals; the last such reversal happened about 780,000 years ago. (Right now Earth’s magnetic north is technically in the Southern Hemisphere.) The direction and angle of lines of force curving between the poles become imprinted in molten rock as it solidifies near the planet’s surface, providing clues to where ancient rocks have been.
To find such traces, the team analyzed rock samples from remote parts of Western Australia and South Africa. These regions contain some of the planet’s oldest cratons, chunks of now stable crust that have survived billions of years of grinding and melting processes and that form the building blocks of continents. The rock layers’ magnetic record shows the shifting of a chunk of the craton in Australia while part of the craton in South Africa stayed still. Such motion is exciting because it “suggests there’s likely to be a plate boundary between the two [cratons],” says Michael Brown, a University of Maryland emeritus geologist, who was not involved in the study.
Multiple researchers agreed that this study’s findings likely represent the earliest we will be able to see such results, as so few rocks remain intact from Earth’s first billion years. “It’s like having a 1,000-piece jigsaw, but you have only 35 pieces,” Brown says. The relative motion doesn’t tell us exactly what was going on in this period, he adds, but it can put new limits on the mathematical models that researchers use to re-create ancient Earth.
The Pilbara Craton in Western Australia holds 3.5-billion-year-old rocks.
Alec Brenner, Harvard University/Yale University
The findings may support another recent study, which uses ancient crystals of the mineral zircon—found in a different part of Western Australia—to suggest that pieces of Earth’s crust may have been melting back into the mantle around 3.35 billion years ago. Evidence from zircon crystals is notoriously difficult to interpret, and the cycling of Earth’s crust into the mantle may happen under many different circumstances. The process is necessary, however, to any form of plate tectonics. In that sense, the two studies reinforce one another.
Fu’s team also found evidence of the earliest known reversal of Earth’s magnetic poles, around 3.46 billion years ago. Together, the study’s findings “demonstrate that Earth was behaving very similarly” to what we see today, according to Jun Korenaga, a Yale University geophysicist, who was not involved in the study.
The Western Australian craton the team studied is home to the oldest confirmed structures left behind by single-celled organisms, which date back roughly 3.48 billion years. Knowing the latitude of those rocks at the time could help researchers learn more about the conditions where early life flourished. And understanding what kind of tectonics operated back then may set limits on how Earth’s modern plate tectonics got started. If we know what Earth’s early tectonics looked like, we can start to hunt for similar behavior on other planets, too.
“What kind of planet did life first appear on?” Fu wonders. The answer, he says, “has implications for how abundant life is likely to be in the universe.”
A version of this article entitled “Early Plates” was adapted for inclusion in the June 2026 issue of Scientific American. This text reflects that version, with the addition of some material that was abridged for print.
