Mato just raised pre-seed Read more

The Day the Plates Started Moving

  • Nov 11, 2026
  • 21 min

Show notes

What the episode covers

In the Paleoarchean, roughly 3,480 million years ago, Earth was a volcanic world of dark rock and open ocean, with no land plants and a crust whose behavior is still disputed.

This episode asks when Earth's crust began moving as plate tectonics: as early as 4,400 million years ago, or only within the last 1,000 million years? Amara and Ola compare a 2026 Science paper on magnetized Pilbara rocks, a Caltech PNAS study of Jack Hills zircons, and a Nature paper that reads the same zircons as evidence of a stagnant lid. They also explain why magnetic records are so fragile and why any start date depends on how started is defined.

It is for listeners who want to understand why the evidence for early tectonics remains contested and what each line of evidence can and cannot show.

Timeline

In this episode

7 moments worth skipping to. The timecodes match the player above.

  1. 0:15Introduction
  2. 1:58Exhibit A: Magnetized Rocks in the Pilbara
  3. 5:07How Fragile Is a Magnetic Memory?
  4. 7:40Exhibit B: Zircons and the Chemistry of Underthrusting
  5. 11:50The Turn: Same Crystals, Opposite Verdict
  6. 15:39A Light Switch or a Slow Creep?
  7. 19:03Outro

Quick answers

Straight from the episode

The questions this one settles, without the listen.

When did plate tectonics start on Earth?
There is no agreed date. Estimates in the episode range from about 4.4 billion to 1 billion years ago. Modern-style tectonics is agreed to have been operating since about 1 billion years ago, but how and when it began is still disputed. Any date also depends on how you define 'started'.
What did the Pilbara magnetized rocks show about early plate motion?
A March 2026 Science paper by Brenner, Fu and colleagues found Pilbara rocks moving poleward at about 47 cm per year while Barberton stayed stationary. That adds up to roughly 2,500 km of drift. The authors caution that both an active lid and an episodically mobile lid fit the data, so it is not a start date for plate tectonics.
Why is it hard to find ancient magnetic evidence for plate tectonics?
Magnetic records in rock are tenuous, and heat and pressure destroy them. Most older rocks have been wiped. The oldest direct evidence is therefore a floor, the earliest record we can still read, and motion may have begun earlier.
How do zircon crystals help scientists study early Earth tectonics?
No earliest rocks remain, but zircons are resilient crystals that survive. A Caltech PNAS study used XANES oxybarometry, U-Pb dating and trace elements. Grain rims at about 3.35 billion years old showed both low and high temperature-to-pressure signatures, consistent with underthrusting. The authors argue for a mobile lid by the early Archean at the latest.
Do all zircon studies agree that plate tectonics started early?
No. A Nature paper on Jack Hills zircons and related zircon magnetism, spanning 3.9 to 3.3 billion years ago, found unvarying latitudes. It argues for a stagnant lid. The disagreement is over what these fragile crystals mean, not over whether the crystals exist.
What is the difference between a stagnant lid and a mobile lid, and is there a compromise?
A stagnant lid is a single immobile shell, while a mobile lid is broken into plates that move and sink. A Journal of the Geological Society paper describes an unstable stagnant lid until the mid-Paleoproterozoic and a gradual transition by the Neoproterozoic. It also offers the compromise term 'chaotic early lid'. The hosts lean differently, Amara toward messy, episodic mobility and Ola toward a slow creep, and no winner is declared.

Transcript

The full conversation

Every word of the episode, 3,064 of them, in the order they were said.

Read the transcriptHide the transcript

OlaPicture it: a sky the color of old mustard, oceans, black rock, and not one tree anywhere.

AmaraOh, man, and volcanoes everywhere, like a planet with a bad temper and no manners.

OlaAnd somewhere under all that, the crust is doing something. Hold on. Let me check something. I want to make sure I'm not overselling the mustard sky.

AmaraHe's fact-checking mid-sentence again. Welcome to Fault Lines. I'm Amara.

OlaAnd I'm Ola. The sky color is artistic license, by the way. Here's the real question. Was that crust one solid shell, a lid, or did it break into plates and start moving?

AmaraAnd did it switch on like a light or creep in through a stagnant lid first?

OlaBecause the answers span a ridiculous range. One camp says about four point forty-four billion years ago. Another says only the last one billion.

AmaraWait, that's a three-billion-year gap. That's not an argument. That's a different planet.

OlaAnd in between, a 2026 science paper lands at about three point Forty-eight billion.

AmaraPlus a Caltech study in PNAS on Jack Hills zircons saying at least three point thirty-five billion.

OlaTiny crystals versus magnetized rocks versus skeptics.

AmaraNo guests, no tricks, just the evidence in a ring and the two of us refereeing.

OlaPlot twist, nobody agrees.

AmaraI love it. Where do we start?

OlaFirst, the rocks that got everyone talking, a paper claiming the oldest direct evidence of motion. Oh.

AmaraOkay, so get this. A chunk of Australian crust called the East Pilbara Craton, and it was apparently on the move, forty-seven centimeters a year, heading toward a pole.

OlaForty-seven centimeters. That's roughly how fast your fingernails grow, give or take. Not exactly a sprint.

AmaraFingernail speed, but that's in the ballpark of modern plates. And here's the kicker. Another chunk, the Barberton Greenstone Belt in South Africa, just sat at low latitudes.

OlaSo one block travels, one block stays put.

AmaraExactly. Relative motion. That's the whole point.

OlaHang on. Who's saying this?

AmaraA science paper from March 2026, Brenner, Fu, and colleagues. It's the one the episode keeps circling.

OlaFine, but you keep calling it direct evidence. Direct to what? Nobody watched a continent drift. They measured magnetism locked in old rock.

AmaraOkay, fair. Direct in the sense that the rock itself remembers where it was. Iron minerals lock in the direction of the magnetic field when the rock forms, and that direction depends on latitude. Zircon chemistry is a hint about conditions. This is a position stamp.

OlaA postmark, basically.

AmaraA postmark from three and a half billion years ago with excitement, and the postmarks disagree between two places.

OlaHow far did the Pilbara travel, though? Forty-seven centimeters a year doesn't mean much until you multiply it.

AmaraScience News did that math, about 2,500 kilometers poleward over several million years, starting around three point four eight billion.

OlaTwenty-five hundred kilometers. That's Oslo to Gibraltar and a bit more.

AmaraPlot it on a map, and it's a serious road trip for a piece of crust.

OlaAnd I'd guess the same team has done this before.

AmaraOh, they have. Science News also points to their earlier work, showing more than 5,000 kilometers of drift starting at three point three four billion.

OlaSo the newer paper pushes the clock back, and the older one shows bigger travel.

AmaraWhich is why I'm a fan. It's not one lucky measurement.

OlaHmm. Now the caveat, because I did read the thing. The authors themselves say this fits rapid collisional active lid tectonics or an episodically mobile lithosphere.

AmaraUgh, you had to bring that up. So even my champion paper doesn't say the switch flipped.

OlaRight. It says something moved, and that's a lot, but it's not a start date.

AmaraI'll allow it, but there's a bigger worry lurking behind all of this.

OlaRight. Magnetism's a delicate thing to build a case on.

AmaraSo how does anyone trust a memory that delicate?

OlaThat's the next question. And after that, we leave the magnets and bring in a very different witness, tiny crystals.

AmaraCrystals on the stand. I love it.

OlaOkay, but Amara, I keep fact-checking myself on one thing. Magnetism in a rock is fragile. Science News quotes Fu calling it a tenuous property.

AmaraTenuous, like a Wi-Fi signal in a basement.

OlaRoughly, yes. Heat and pressure destroy it easily, and a rock that's been buried, cooked, and squeezed for billions of years has plenty of both.

AmaraSo the surprise isn't that the signal is faint.

OlaIt's that any of it survived at all. Science News says for rocks older than the Pilbara ones, most have simply lost their magnetic record.

AmaraWiped.

OlaWiped clean. Think of a receipt in the bottom of a coat pocket. Run it through the wash a few times, and the ink goes.

AmaraAnd you're holding the one receipt that still has a date on it.

OlaExactly. And the stamp says three and a half billion, but you don't know the purchase happened that day.

AmaraWait. Why not? It's stamped right there.

OlaBecause the stamp is only the oldest one that survived. Everything older went through the wash and came out blank.

AmaraOh, that's sneaky. So a blank pile of rocks doesn't mean nothing was moving.

OlaRight. Science News includes a view that some motion may have begun even earlier than that.

AmaraHuh.

OlaSo the oldest direct evidence is a floor. Nothing we can read says the plates started moving before it, but nothing says they didn't.

AmaraOkay, I love this because it flips how I was hearing the headline. Oldest evidence sounds like a birth certificate.

OlaAnd it's closer to the earliest photo of a person you found in an attic. They were alive before the photo.

AmaraFair, though somebody's gonna say if the earlier rocks are blank, you're just guessing, pure vibes.

OlaAnd that's a fair jab. A blank receipt isn't proof of a purchase, but it isn't proof of no purchase either. It's just silent.

AmaraSo what do you do with a silence?

OlaYou respect it. You say, "This is how far we can read, and past that line, we're in the dark, not in some different story."

AmaraThere's a practical upshot for anyone scrolling headlines too. When you see earliest evidence, hear it as earliest we can still read.

OlaWhich is why geologists want a second witness that doesn't rely on magnets at all, something that doesn't get wiped the same way.

AmaraOh, you're gonna love this part. We have exactly that.

OlaCrystals.

AmaraTiny ones. Now, the zircons say something similar in a completely different language, and they're not shy about it.

OlaZircon, before we trust a crystal the size of a grain of sand, why zircons? Why not just read the rocks?

AmaraBecause the rocks are gone. Physorg covering the Caltech work in March 2026 says no rock records from Earth's earliest years remain.

OlaNone?

AmaraNone. So researchers turn to zircons, which resist physical and chemical alteration over time. They keep a precise chemical record of the moment they formed.

OlaA time capsule that survives the demolition of the building.

AmaraOkay, yes. That's better than mine. So the Caltech team, the PNAS paper, Hutcheon, Tissot, and colleagues used Jack Hills zircons and threw three tools at them.

OlaWhich three? I like knowing what's in the toolbox.

AmaraRedox oxidation state measurements called Zain's oxybarometry, uranium-lead dating, and trace elements.

OlaAnd the crystals have rims, right? Layers added later, like tree rings.

AmaraExactly. And the rims at about three point three five billion years old show two signatures, low and high temperature to pressure ratios.

OlaBoth in the same set?

AmaraBoth. That mix is typical of plate underthrusting. One slab of crust going under another gives you cold high-pressure conditions and hot ones.

OlaWalk me through that because temperature to pressure ratio is a mouthful.

AmaraThink of how hot the rock is for how deep it's buried. A low ratio means squeezed but relatively cool, which is what you'd expect if a cold slab is being shoved downward.

OlaA high ratio means hot for its depth.

AmaraSo the low ratio is the fingerprint of something being pushed down.

OlaThat's the reading, and it's why getting both signatures in one set of rims matters. One signature alone is a lot easier to explain away.

AmaraHold on. What did the older rims show?

OlaPlot twist. The Hadean rims show only the high ratio, one flavor.

AmaraHuh. So the older crystals have one story, and the younger ones have two.

OlaAnd the authors read that as a change. They say mobile lid tectonics was active by the early Archean at the latest.

AmaraAt the latest. That phrase is doing careful work.

OlaRight? They're giving a ceiling, not a birthday. They also say the mantle had near modern redox states by about four point fifteen billion years ago.

AmaraRedox, meaning oxidation state of the mantle, which is a different question from whether plates moved.

OlaDifferent question, same crystals. Let me poke at that anyway. Why bring up the mantle if it's a different question?

AmaraMy read, not theirs, it's context. If the deep Earth already looked chemically modern that early, the crystals aren't coming from some alien planet. That doesn't prove plates. It just makes the plate reading less exotic.

OlaAnd for anyone just joining, remind me what mobile lid means since we'll lean on it.

AmaraA mobile lid is a surface broken into pieces that move. A stagnant lid is one rigid shell. Keep that pair in your pocket.

OlaOne more flag before I get ahead of myself. These are rims on tiny grains, so I take this as a pattern in the chemistry, not a photograph of plates.

AmaraAgreed. It's a clue about conditions, and at the latest is the author's own humility about it.

OlaOkay. My notebook says two independent lines now. Magnetic rock in one corner, zircon chemistry in the other. Are we done?

AmaraOh, Ola, you want a tidy ending.

OlaI want to know whether anyone's checked the crystals the way I'd check a witness.

AmaraSomebody has, and they looked at the very same crystals and heard a completely different story.

OlaSame crystals?

AmaraSame crystals.

OlaYes, there's skeptics. A peer-reviewed nature paper titled Hadean to Paleoarchean Stagnant Lid Tectonics Revealed by Zircon Magnetism. Magnetic signals in the title Jack Hills zircons and related grains spanning about three point nine to three point three billion years imply unvarying latitudes.

AmaraUnvarying, as in the crystals never went anywhere?

OlaRight. The authors read that as stagnant lid behavior. One rigid shell, no plate tectonics

AmaraSo walk me through the logic, though. Why does staying at one latitude mean no plates?

OlaBecause magnetism records latitude. If plates were carrying crust around, you'd expect those latitudes to wander over time. If the signals keep pointing to the same latitudes across that whole stretch, the simplest reading is that nothing was drifting.

AmaraSo the paper is basically saying the crystals stood still.

OlaThat's the argument. Simplest reading is my phrase, not theirs.

AmaraHold on. Same Jack Hills zircons as the Caltech world?

OlaCareful. Zircons from the same Jack Hills region, yes. I can't tell you the two teams measured the very same grains.

AmaraOh, man, so one region, two teams, and one says subduction while the other says the lid never cracked.

OlaAnd the window overlaps. The nature range runs right across the three point three five billion mark the Caltech rims sit on.

AmaraWhich means it collides head-on with the Caltech claim and with the Pilbara motion at three point four eight.

OlaTwo, three papers, one stretch of time, and they can't all be right in the simplest reading.

AmaraWait for it, though. The evidence types differ. Pilbara is magnetized rock. Caltech is chemistry, and this one is magnetism inside the crystals.

OlaWhich is what bothers me. Magnetism is on both sides of this fight, so you can't sort it out by method alone.

AmaraOkay, I'll defend the chemists. Rims with both low and high T over P look like Underthrusting. That's a pattern you'd expect from plates sliding under each other. Show me a stagnant lid doing that.

OlaI'll defend the lid then, and this part is our gloss, not the paper's. The word detrital means loose grains eroded out of older rock and carried off. A grain with that history is a long way from a clean recording.

AmaraThen here's my glass back. Maybe the camps aren't measuring the same moment. Rims grew late on grains that were already old. A rim can speak for one place in time while the grain's magnetism speaks for another.

OlaUh, that would let both be right, but unvarying is a big word. To hold up, those latitudes would have to stay put across the whole three point nine to three point three stretch, and one wandering grain breaks the claim.

AmaraMm. So the fight isn't old versus young at all.

OlaIt's the same fragile crystals actually mean. Nobody here disputes that the crystals exist. The argument is over their reading.

AmaraWhich answers your question from a minute ago, two independent lines of evidence. So are we done? Not even close.

OlaNot even close. Two lines pointing one way is nice. A third line from zircons out of the same region pointing the other way puts everything back on the table.

AmaraAnd that's the part I did not see coming. I walked in thinking someone was going to be wrong about a date.

OlaMe too. But if both camps can read the crystals and get opposite answers, maybe a single start date is the wrong frame.

AmaraOoh, then you need a third camp, the ones who say plates crept in a bit at a time.

OlaThe creepers. Okay, so the creepers. A Journal of the Geological Society paper says Earth was an unstable, stagnant lid planet until the mid-Paleoproterozoic.

AmaraMid-Paleoproterozoic, so that's way after the zircon people say things were already moving.

OlaRight, and the transition afterwards is gradual, finishing by the Neoproterozoic. No single morning.

AmaraHm, a dimmer switch. Somebody turned it up over a billion years.

OlaHm, and notice the word unstable. It isn't saying the lid was frozen solid. It's saying it kept misfiring.

AmaraMisfiring how?

OlaThe paper doesn't hand me a mechanism in what I have. It just says unstable, and it gives you a long, bumpy handover.

AmaraFine. Does it at least say where everybody agrees?

OlaIt does. The Journal of the Geological Society paper says all sides except modern-style plate tectonics has run since about one billion years ago.

AmaraSo that's the floor of the floor.

OlaHeh, sure, but before that, it says there's no consensus on when it began or what came before.

AmaraOh, man. That's a rough sentence for a podcast called The Day the Plates Started Moving.

OlaHeh, we did title it with a date in mind.

AmaraWait for it, though. It also surveys the Hadean onset claims, right? The really early crowd?

OlaIt does, and it offers a compromise term: chaotic early lid.

AmaraChaotic early lid. Heh, that sounds like my hair in the morning.

OlaHeh, it's a diplomatic phrase. Neither side has to say the word stagnant or the word plate.

AmaraHere's what I like about it, though. Chaotic early lid sits between our two clean pictures. Not one rigid shell, not modern plates, something lurching in between.

OlaAnd it lets everyone keep a seat at the table. The earliest claim we opened with about four point four billion years ago sits at one end of the range, and the one billion line sits at the other.

AmaraWhich is why I keep saying a single date is a trap. Whichever date you pick, you're also picking a definition of what counts as started.

OlaRight. Does a few rocks drifting count? Does one slab sliding under another count? Or do you need a planet-wide jigsaw?

AmaraOkay, lean time. Mine is messy. The Pilbara authors themselves said episodically mobile fits their data, and I keep coming back to that.

OlaEpisodically mobile

Speaker 3Pieces moving, then stopping.

AmaraExactly. A rock that drifts for several million years and a rock that sits still can both be true on the same planet at different spots in different eras.

Speaker 3Yeah. That's the thing about a handful of samples. There are a few dots on a map that's mostly missing.

AmaraAnd dots can be perfectly real without telling you the shape of the picture.

Speaker 3And the chemistry?

AmaraChemistry says something was colliding or sliding under at some point. It doesn't say how much of the globe was doing it.

Speaker 3So the switch camp can't prove a switch?

AmaraAnd the creep camp can't prove a creep. Everyone's holding a few crystals and a few rocks.

Speaker 3Which is humbling, honestly. Three and a half billion years, and our evidence fits in a shoebox.

AmaraA small shoebox. Okay, time to land this and say what a listener can actually carry out the door.

Speaker 3Ooh, yes. What do you keep? So where do I land after all this? Creep. Slow, messy, with a lid that sulked for a long time before anything moved. I keep thinking of Oslo winters, where nobody can name the day the snow stopped being snow.

AmaraThat is a very Norwegian way to dodge a date.

Speaker 3It's a lean, not a dodge.

AmaraFine. Mine's lumpy. Fits and starts, plates waking up, napping again. Nobody gets a birthday, and I'm weirdly at peace with that.

Speaker 3And we actually disagree. That wasn't staged for the microphone.

AmaraNot even a little. What I take from it is how you weigh evidence this old. A few tiny crystals, a few magnetized rocks, every one fragile and every one open to opposite readings.

Speaker 3So you hold the dates loosely and watch for the next crystal.

AmaraAnd you ask what survived. The oldest evidence is a floor, not a birth certificate, and the silence below it isn't an answer either way.

Speaker 3Plus, you remember the same crystals can carry opposite verdicts, so the question is never just how old, it's who's reading them.

AmaraExactly. Now, the homework. Subscribe, leave a review, and email us at hello@heymaddo.com.

Speaker 3Tell us your favorite epoch or the theory you think we got wrong.

AmaraOh, please say we got something wrong. I wanna be argued with.

Speaker 3Me too. I keep a notebook for exactly that.

AmaraOf course you do.

More episodes

Keep listening

Other episodes of Fault Lines, newest first.

All episodes of Fault Lines

Sources

Where this came from

6 reports behind the episode. Every one of them opens where it was published.