Isn't motion relative? Why doesn't the symmetry work both ways?

Relative motion at constant velocity is symmetric — each observer sees the other's clock as slow. The twin paradox involves the traveler ACCELERATING to turn around. Acceleration is not symmetric: only one twin's frame undergoes it. That asymmetry resolves the paradox: special relativity treats inertial (constant-velocity) frames equally, but acceleration breaks the equality.

Has this been measured?

Yes, with atomic clocks. The Hafele–Keating experiment (1971) flew cesium clocks around the world in both directions and compared them to ground clocks — the results matched relativity's predictions including the asymmetric aging. Many follow-up experiments since have confirmed it down to nanosecond precision. The astronaut Sergei Krikalev has accumulated about 20 milliseconds of asymmetric aging from his orbital stays.

What if the traveler doesn't accelerate (no turnaround)?

Then they don't come back to compare clocks. Each twin sees the other's clock as running slow — and both views are consistent, because they're separated. The paradox needs a reunion, which requires at least one twin to change inertial frames.

The Twin Paradox — Why Both Twins Don't Age the Same

The setup

Two identical twins. Twin A stays on Earth. Twin B climbs into a rocket and flies to a star 10 light-years away at 80% the speed of light, then turns around and flies back.

How much time passes for each twin?

The relativistic factor at 0.8c is γ = 1/√(1−0.64) = 1.67. So:

Twin A's perspective (Earth frame): Twin B's clock runs slow by a factor of 1.67. The round trip takes Twin A about 25 years (10 light-years out + 10 back, at 0.8c). Twin B, whose clock ran slow, aged 25/1.67 ≈ 15 years during the trip.
Twin B's perspective (rocket frame, except during turnaround): Twin A's clock runs slow during the cruise legs by the same factor of 1.67. Twin B sees Twin A's clock ticking only 60% as fast as their own during the trip.

When they reunite, Twin B has aged 15 years and Twin A has aged 25 years. Twin B is younger.

The paradox is the apparent symmetry. Each twin, during constant-velocity cruise, sees the other twin's clock as the slow one. So how do they end up with the traveler aging less and not the stay-at-home?

The resolution

The two situations are not symmetric. Twin A stays in one inertial frame the entire trip. Twin B changes inertial frames at the turnaround — accelerating, decelerating, and re-accelerating in the opposite direction.

Acceleration is detectable. Twin B feels themselves being pushed back into their seat at the turnaround; Twin A doesn't. That's a real physical difference, and special relativity treats it differently.

The cleanest way to see what happens: during the turnaround, when Twin B is changing inertial frames, their plane of simultaneity (the set of distant events they call "now") sweeps forward through Earth's history. From Twin B's perspective, Earth's clock "jumps forward" rapidly during the turnaround. That jump accounts for the extra aging Twin A accumulates relative to Twin B.

Once the trip is over and both are back in the same inertial frame at the same location, the asymmetry is unambiguous: Twin B's clock has accumulated less time.

Where the symmetry actually breaks

The full statement: special relativity says inertial frames are equivalent for the laws of physics, but acceleration is not invariant. A frame undergoing acceleration knows it's accelerating (you can feel the force, you can drop a ball and see it doesn't fall straight). An inertial frame can't be distinguished from another inertial frame in motion relative to it.

In the twin paradox, only Twin B undergoes acceleration. The turnaround is what marks one twin as the "traveler" in a physically meaningful sense.

If both twins fired their engines and accelerated symmetrically (one going to the star and the other going elsewhere), and both came back, neither would be the "younger" one — they'd age identically. The asymmetry comes from one twin having a non-inertial portion of their worldline and the other not.

The spacetime-diagram picture

Plot time on the vertical axis and space on the horizontal. Twin A's worldline is a vertical line — they stay at the same place, just moving forward in time.

Twin B's worldline is a triangle — they go diagonally up and to the right (cruise out), turn at the top, and go diagonally up and to the left (cruise back), reuniting with Twin A's line.

The "length" of a worldline in spacetime — measured with the relativistic interval √(t² − x²/c²) — is the proper time experienced by the traveler. Vertical worldlines are longer (more proper time accumulated). Diagonal worldlines are shorter. Counterintuitively, the longer-looking path through space gives less proper time.

Twin A's worldline is a straight vertical: maximum proper time. Twin B's worldline is a triangle: shorter proper time. So Twin A ages more, Twin B ages less. The geometry of spacetime is what makes this work.

Is this really observed?

Yes, repeatedly, at multiple scales.

Hafele–Keating (1971). Atomic clocks were flown on commercial jets around the world, eastbound and westbound. The eastbound clock (moving with Earth's rotation) lost time; the westbound clock (moving against rotation) gained time. Both differences matched general relativity to a few percent. The paper is a classic.

GPS satellites. Their atomic clocks have to be corrected for both special and general relativistic effects daily. Without correction, GPS would drift by km/day. The corrections embed the same physics as the twin paradox.

Particle accelerators. Muons stored in storage rings at near-light-speed live much longer (lab frame) than at-rest muons would. The lifetime extension exactly matches the relativistic gamma factor.

Astronauts. Sergei Krikalev's 803 cumulative days in orbit have left him roughly 20 milliseconds younger than he would have been on the ground. Small, but measurable in principle and predicted by relativity.

The twin paradox is not a thought experiment that's "kind of true" — it's a precise prediction of special relativity that has been confirmed many times with various clocks.

A variation: triple paradox

Suppose Twin A stays on Earth, Twin B goes to the star, and Twin C goes to a different star in another direction, both at 0.8c, both turning around and returning at the same time. When all three reunite, A is oldest, and B and C are equally younger.

The asymmetry isn't about "which twin moved." It's about which twins underwent acceleration. The accelerated twins age less, regardless of direction. Twin A stays inertial; B and C both leave the inertial frame symmetrically.

What it isn't

It isn't about who "thinks" they're moving. Motion is relative, but acceleration isn't. The asymmetry is physical.

It doesn't depend on the details of the turnaround. You can make it instantaneous (mathematical idealization) or gradual (more realistic) — the total age difference comes out the same as long as Twin B comes back to Twin A.

It isn't paradoxical, in the end. "Paradox" here means "apparently contradictory but actually consistent." Once you account for the asymmetry of acceleration, everything checks out. Special relativity has been internally consistent for 120 years.

The takeaway

The twin paradox is the cleanest dramatization of time dilation: take two identical clocks, send one on an accelerated journey and keep the other stationary, and they will not read the same time when reunited. The "paradox" is just the symmetric-looking situation falling apart under closer inspection: only one twin actually undergoes acceleration, and that one ages less. The effect is real, measured, and embedded in technologies we use every day.