How Fermilab Proved Einstein's Speed of Light

Here's why I — a biotech reporter — ended up deep in a physics rabbit hole at 11pm: particle accelerators.

The same machines that smash protons to confirm Einstein are also the ones being miniaturized for cancer therapy. Proton beam treatments work because of relativistic physics — you can't accurately target a tumor with a particle beam if you don't account for the fact that fast-moving particles behave differently than Newton predicted. So when Don Lincoln, a particle physicist at Fermilab, sat down with Lex Fridman to explain how we actually know the speed of light is constant, I had a professional reason to pay attention. And then the science just got interesting on its own terms. 🧬 (adjacent, I know — bear with me.)

Lincoln's explanation cuts straight to what makes special relativity so stubbornly hard to accept: it's not that the math is complicated. It's that the central claim sounds insane.

The assumption that broke physics (in a good way)

Einstein's 1905 theory rested on two premises. The first — that the laws of physics work the same for everyone regardless of how they're moving — is old news. Galileo had that one. Newton had that one. If you've ever tossed a ball on a moving train, you intuitively get it.

The second premise is the one that causes problems: everybody measures the speed of light as the same number, regardless of how fast they're moving relative to each other. You, stationary on a platform. Me, hurtling past at 90% the speed of light. We both clock light at 299,792,458 meters per second. Not approximately the same. The same.

That's the part that made physicists in 1905 do a double-take.

Because classically, speeds add. A ball thrown forward from a moving train travels faster than one thrown from a standstill. That's just arithmetic. But light? Light does not care about your train. Lincoln puts it plainly: the constancy of light speed "is very different from what Newton would have said or Galileo or any of the old guys." Combine that with the first premise — everyone's reference frame is equally valid — and you get time dilation, length contraction, E=mc², and about a century of physics students quietly suffering.

A quick note for anyone who's seen the Michelson-Morley TikToks: yes, that 1887 experiment had already produced a null result strongly suggesting light doesn't pick up speed from a moving source. Whether Einstein consciously built on that finding when he wrote his 1905 paper is genuinely debated among historians of science — he may have arrived at the constancy principle through his own reasoning about electromagnetism rather than directly from Michelson-Morley. Lincoln frames Einstein's starting point as a premise, not a derivation from prior experiment. Both things can be true simultaneously, and that ambiguity is actually kind of interesting.

Okay but how do we know it's true

Lincoln is good at this part. He doesn't hand-wave. He describes a real class of experiment — one that wasn't available to Einstein but is now routine at facilities like Fermilab.

Some subatomic particles decay by emitting photons. So: smash things together in a collider. A stationary particle decays, emits a photon, photon travels to the detector. Time it. It's c. Obviously — that's the baseline.

Now here's the test. Sometimes those collisions produce fast-moving particles — some traveling at 95%, 97% of the speed of light. Those particles also decay into photons. Classical physics says the photon should inherit the parent particle's velocity on top of its own: roughly 1.97c, arriving at the detector measurably faster than light from a stationary source. It doesn't happen. The photon still clocks in at c. Every time.

(A precision note: the "half the time" framing you might hear is a shortcut that assumes a specific geometry — the real point is that classical velocity addition predicts a meaningfully faster arrival that simply doesn't materialize.)

This, Lincoln says, "is a hard, serious measurement." Not a thought experiment. Not a philosophical argument. An actual number that comes out wrong if Einstein was wrong, and keeps coming out right.

The part where I put on my ethics hat for a second

The evidence for c being constant is genuinely overwhelming — I'm not surfacing doubt where there isn't any. But I do find it philosophically interesting that since 1983, the speed of light isn't actually measured anymore. The meter is now defined in terms of c. We locked in 299,792,458 m/s and built the unit of length around it. Which means if someone ran Lincoln's experiment tomorrow and got a slightly different number, we'd conclude the detector was miscalibrated, not that Einstein was wrong.

Is that circular? Sort of. Is it bad science? No — it reflects how confident the entire edifice of modern physics is in this value. But it does mean we're in a regime where "testing" the constant speed of light has a specific, subtle meaning that's worth being clear about. We're testing the predictions that follow from it. The constant itself has graduated from hypothesis to definition.

Why does Fermilab's measurement specifically earn our trust? Because particle accelerators have to get relativistic physics right to function at all. The engineering stakes are a built-in audit.

Minkowski and the upgrade nobody talks about

Three years after Einstein's 1905 paper, mathematician Hermann Minkowski gave a lecture — published the following year in 1909 — that reframed the whole thing. Not as a theory about the behavior of light and moving objects, but as a theory about the geometry of space-time itself. Space and time aren't separate containers that happen to interact weirdly. They're coordinates in a single four-dimensional structure. Light's speed isn't a rule imposed on top of space-time; it's a feature of space-time's geometry.

Lincoln describes this as a kind of cognitive unlock: "The thing to remember is the speed of light — it's the speed of light through space-time. Once you embrace that, that makes a whole ton of sense."

And I actually believe him that it helps, even if I can't fully feel it intuitively. The analogy he reaches for: c is a property of space the same way space has a maximum electric field strength it can support. Whatever space is — and Lincoln is refreshingly honest that "we don't know what space is" — it has this one speed built into it. Everything else follows from refusing to unify space and time, which is apparently what our senses insist on doing and what Newton obliged.

Where it gets weird again

Here's the open question I'd actually want you to Google after reading this: Is c truly constant everywhere in the universe, or just locally?

There are serious physicists — a minority, but not cranks — who study whether fundamental constants might vary across cosmic time or space. The fine-structure constant, which involves c, has been examined using light from ancient quasars. Most measurements find no variation. Some find hints of something. The field is alive.

Lincoln's framing — c as a property of space itself — implies it's universal and unchanging. That's the overwhelming scientific consensus. But "the universe might not have the same physics everywhere" is an open research question in a way that "Einstein was probably wrong about this specific prediction" no longer is.

The particle physicists at Fermilab have done their part. Now the cosmologists are running the longer experiment.

— Mei Zhang covers biotechnology, genetics, and the future of medicine for Buzzrag.