University Physics IV · Special Relativity II · 3.3

The Twin Paradox

Mutual time dilation is not a contradiction, but you have to show that. This lesson gives you three independent ways to settle who is younger at the reunion — a worldline length, a simultaneity slice that jumps, and a count of light pulses — and why the traveller's g-force is not the reason.

Open the lab ↓Practice & worked examples ↓

Build the model

Connect the measurement to the mechanism.

Time dilation is symmetric between inertial frames: while Alice cruises away at 0.80c, each twin correctly measures the other's clock running at 1/γ = 0.60 of their own rate. That symmetry is not the paradox. The paradox is that the twins meet a second time, and a reunion is a single event about which no frame is allowed to disagree.

What breaks the tie is that only Bob keeps one inertial frame from departure to reunion. Alice's worldline is bent, and Minkowski geometry carries a minus sign — dτ² = dt² − dx²/c² — so of all worldlines joining two timelike-separated events, the straight one accumulates the most proper time. Bob's 10.0 years against Alice's 6.0 is the triangle inequality run backwards, and nothing more.

What the model costs is that Alice cannot write her own account in a single frame: her present exists only leg by leg, and when she swaps legs the slice she calls now sweeps 6.4 years up Bob's worldline while absolutely nothing happens to his clock. Every consistent story she tells has to carry that jump — or drop slices entirely and count the light pulses that actually arrive, which is the one piece of bookkeeping no frame can dispute.

Simple definition: The twin paradox is the apparent contradiction that two twins each measure the other's clock as slow, resolved by the fact that only the twin who stays inertial has a straight worldline, and the straight worldline between two meetings carries the most proper time.
Example: Alice flies 4.0 ly out at 0.80c and straight back while Bob waits: Bob's straight worldline records 2 × 4.0/0.80 = 10.0 years, Alice's bent one records 10.0/γ = 10.0 × 0.60 = 6.0 years, so she returns 4.0 years younger.

Proper time of one inertial legΔτ = √(Δt² − (Δx/c)²) = Δt/γ

The minus sign is the whole story: adding displacement subtracts proper time, so a detour costs ageing instead of buying it.

Δt and Δx are the leg's time and displacement in one inertial frame; Δτ is what the clock carried along that leg reads.

Bent worldline against straightτ = T/γ = 2D/(γv), Δage = T(1 − 1/γ)

One division settles the reunion: D = 4.0 ly at β = 0.80 gives T = 10.0 yr, τ = 6.0 yr, and a 4.0 yr gap.

D is the one-way distance in the home frame, T = 2D/v the home elapsed time, v the constant cruise speed on each leg.

Bondi k, the Doppler factork = √((1 + β)/(1 − β))received rate = k or 1/k

Converts clock rates into what a twin can actually see through a telescope, with the light travel time already folded in.

β = v/c, dimensionless. An approaching twin receives at k, which is 3 at β = 0.80; a receding twin receives at 1/k, which is 1/3.

The signal ledger closes(1/k)(D/v + D/c) + k(D/v − D/c) = 2D/(γv) = τ

Arriving pulses are events, so this route to the age gap never chooses a simultaneity convention at all.

Left: Bob's years of watching her recede, then approach, each weighted by the rate he actually sees. Right: her proper time.

Simultaneity jump at turnaroundΔtⱼᵤₘₚ = 2vD/c² = 2βD/c

It grows with the distance to the star and with β, and it delivers no signal, no force and no energy to Bob.

D is the turnaround distance in Bob's frame; at β = 0.80 and D = 4.0 ly the jump is 6.4 yr of Bob's time.

Reciprocity is real; it is not the paradox

While both twins move inertially, each is entitled to say the other's clock runs slow by the same factor — at β = 0.80 each says the other ticks at 0.60 of their own rate — and both are right. There is no contradiction, because the two statements are not about the same pair of events. Each twin is comparing one clock of their own against a pair of separated clocks in the other frame, and those pairs are synchronised differently. The paradox only bites when the twins meet a second time, because who is older at the reunion is decided at one place at one moment, and no frame can dodge it. So the question is never which twin's dilation is real. It is which twin's account of the whole journey can be written in a single inertial frame.

Proper time is worldline length, and straight is longest

Strip the frames out and the problem is geometry. Along any worldline dτ² = dt² − dx²/c², and the elapsed proper time is the integral of dτ — the Minkowski length of the path between two events. Bob never moves in his own frame, so dx = 0 and his clock records the coordinate time itself, 10.0 yr. Alice's outbound leg covers Δt = 5.0 yr and Δx = 4.0 ly, so Δτ = √(5.0² − 4.0²) = √9 = 3.0 yr, and the return leg gives another 3.0 yr: 6.0 yr in all. The minus sign is what makes this unlike Euclidean geometry. There the straight line is the shortest path; here the straight timelike worldline between two events is the one with the most proper time, and every bend costs ageing. Alice returns younger because her path through spacetime was bent — a statement that names no frame at all.

The traveller's books, and the slice that jumps

Alice may still tell the story from her side, provided she admits she is using two frames. On the outbound leg she computes Bob's clock at 1/γ of hers: over her 3.0 yr he ages 3.0 × 0.60 = 1.8 yr, and the inbound leg credits him with another 1.8 yr. That accounts for only 3.6 of his 10.0. The missing 6.4 yr appears the moment she changes frames. In the outbound frame the event on Bob's worldline simultaneous with her turnaround is t = 5.0 − βD/c = 5.0 − 3.2 = 1.8 yr; in the inbound frame the same turnaround is simultaneous with t = 5.0 + 3.2 = 8.2 yr. Her present has swung across 6.4 yr of his life, which is the Δtⱼᵤₘₚ = 2vD/c² of the formula card, and 1.8 + 6.4 + 1.8 = 10.0. Nothing crossed the gap: no signal, no force, no influence on Bob's clock. What changed is which of Bob's birthdays Alice counts as happening now. Give the turnaround a finite duration and the slice sweeps instead of jumping — smoother, and the same 6.4 yr.

Count the pulses: an account with no convention in it

The cleanest resolution never mentions simultaneity. Let each twin send one pulse per year of their own proper time and use the Doppler factor k = √((1 + β)/(1 − β)), which is √(1.80/0.20) = 3 at β = 0.80: a receding observer receives at 1/k, an approaching one at k. Alice's telescope therefore delivers 1 pulse across her outbound 3.0 yr and 9 across her inbound 3.0 yr, and 10 is Bob's age at the reunion. Bob's own tally runs the other way and returns 6, which is Alice's. Every pulse emitted is received, and an arrival is a single event, so those totals carry over unchanged into any frame and no synchronisation convention is ever selected. What separates the twins is when the rate flips: Alice sees 1/k become k at the instant she turns, while the news of that turn has to cross 4.0 ly of space before it reaches Bob. That delay, and not a g-force, is the asymmetry in a form each twin can point at.

The turnaround marks the asymmetry, it does not cause it

Replace Alice with two probes. An outbound probe passes Bob at 0.80c and zeroes its clock; at the star it passes an inbound probe running the other way and hands over its reading of 3.0 yr; the inbound probe adds its own 3.0 yr and calls out 6.0 yr as it passes Bob, whose clock reads 10.0. Nobody in that version ever accelerates, and the answer does not move. What the handoff keeps is the bend: the relay's combined worldline is still not straight. Acceleration is simply how one traveller changes frames, which makes it a reliable marker of who bent — but it is not what sets the size. Δage = T(1 − 1/γ) is built from β and T alone, so the gap tracks how fast she cruised and for how long, and is deaf to how sharply she turned.

The version that has actually been measured

Muons stored in a ring at CERN circulated at γ = 29.3 on a closed loop — out and back, with the laboratory playing the stay-at-home twin. Their proper lifetime of 2.197 µs appeared in the laboratory as 29.3 × 2.197 = 64.4 µs, matching γτ₀ to about one part in a thousand. The circular path means their proper acceleration was around 10¹⁸ g, beyond anything a spacecraft could survive, and it shifted the decay rate by nothing measurable. That is the clock hypothesis in raw form: the rate of an ideal clock depends on its instantaneous speed and not on its acceleration. It is an experimental result rather than a theorem of the two postulates, and it is exactly what licenses treating Alice's turnaround as a corner rather than as a separate physical effect.

Change one variable at a time

Make the relationship visible.

Interactive model

Cruise speed β = v/c0.80 c

Where Alice is in her trip0.45

Nudge Alice's position from 0.45 to 0.50 and the Bob-year she calls now leaps from 1.62 to 8.20 while her own clock ticks on evenly; the bar marks the 6.4 yr skipped by the jump taken at the turnaround itself. Then raise β and watch the bar and the reunion gap grow together.

ALICE'S CLOCK2.70 yr

BOB'S CLOCK ON HER SLICE1.62 yr

AGE GAP AT REUNION4.00 yr

SLICE JUMP AT TURNAROUND6.40 yr

Live interpretationALICE'S CLOCK: 2.70 yr. BOB'S CLOCK ON HER SLICE: 1.62 yr. AGE GAP AT REUNION: 4.00 yr. SLICE JUMP AT TURNAROUND: 6.40 yr

Catch the common trap

Explain before calculating.

Misconception check

The usual account says Alice ages less because she accelerated and Bob did not, so the g-forces did it. Acceleration does break the symmetry, but it does not set the number. Hold the cruise speed and the distance fixed and make the turn ten times fiercer over a tenth of the time, and the gap barely moves; leave the turn exactly as it was and send her to a star twice as far away, and the gap doubles. Nor does the argument need an accelerated observer at all: pass a clock reading from an outbound probe to an inbound one at the star and three inertial observers still deliver 6.0 yr against 10.0 yr. Behind that stands the clock hypothesis, that an ideal clock's rate depends on speed alone, tested to about 10¹⁸ g in muon storage rings. Acceleration reliably identifies the twin who changed frames; the ageing is charged to the bend in her worldline, not to the force that made it.

Alice flies 4.0 ly out at 0.80c and straight back at 0.80c while Bob stays home, so γ = 5/3 and the turnaround is brief. Which statement about their reunion is correct?

Choose an answer to test the model.

Practice & worked examples

Reason from the model, then test the result.

EasyAlice travels to a star 3.0 ly away at 0.60c and returns at the same speed while Bob waits at home. Find how much each twin ages, and check the result against the Newtonian limit.

Work in Bob's frame first, where the bookkeeping is easiest: the round trip covers 6.0 ly at 0.60c, so T = 6.0/0.60 = 10.0 yr on his clock. Because he never moves in his own frame, that coordinate time is also his proper time.
γ = 1/√(1 − 0.60²) = 1/√0.64 = 1/0.80 = 1.25, so 1/γ = 0.80.
Alice's clock runs at 1/γ of Bob's throughout both legs, and the turnaround is brief, so τ = T/γ = 10.0/1.25 = 8.0 yr.
Age gap = 10.0 − 8.0 = 2.0 yr. Check the limit: as β → 0, γ → 1 and the gap → 0, which is the Newtonian answer the model has to reproduce.

AnswerBob ages 10.0 yr, Alice 8.0 yr, so she returns 2.0 yr younger.

MediumAlice flies 24 ly out at 0.96c and straight back at the same speed, so 1/γ = 0.28 and γ = 25/7. Account for every one of Bob's years from Alice's side, using her two inertial frames, and say where the bulk of them arrive.

In Bob's frame the one-way time is 24/0.96 = 25.0 yr, so T = 50.0 yr, and he reads all 50.0 on his own clock because he never moves in his frame.
Alice's proper time on the outbound leg is √(25.0² − 24.0²) = √(625 − 576) = √49 = 7.0 yr, and the return leg matches it, so τ = 14.0 yr. The reunion gap is 50.0 − 14.0 = 36.0 yr, which agrees with T(1 − 1/γ) = 50.0 × 0.72 = 36.0 yr.
Inside the outbound frame Alice credits Bob with 7.0 × 0.28 = 1.96 yr, and the inbound frame credits him with another 1.96 yr. Her two legs together cover only 3.92 of his 50.0 yr.
The other 46.08 yr arrive in the frame change: Δtⱼᵤₘₚ = 2βD/c = 2(0.96)(24) = 46.08 yr. Read it off the two slices — the outbound frame calls t = 25.0 − 0.96 × 24 = 25.0 − 23.04 = 1.96 yr simultaneous with the turnaround, the inbound frame calls t = 25.0 + 23.04 = 48.04 yr, and 48.04 − 1.96 = 46.08 yr. The inbound leg then carries Bob from 48.04 to 50.00 yr.
Sum: 1.96 + 46.08 + 1.96 = 50.00 yr ✓. At this speed the ledger is almost all jump — the two dilated legs supply under 8% of Bob's life, and the rest is bookkeeping about which of his distant birthdays Alice is entitled to call the present.

AnswerBob's 50.0 yr splits as 1.96 + 46.08 + 1.96: two time-dilated legs of 1.96 yr each, plus a 46.08 yr simultaneity jump at the turnaround.

HardBack to the standard trip (β = 0.80, D = 4.0 ly, T = 10.0 yr, τ = 6.0 yr). Each twin transmits one pulse per year of their own proper time. Count the pulses each one receives, show that the totals are the other twin's age, and prove the count reproduces τ = T/γ for any β.

Doppler factor: k = √((1 + 0.80)/(1 − 0.80)) = √(1.80/0.20) = √9 = 3. Receding, a twin receives at 1/k = 1/3 of the emitted rate; approaching, at k = 3.
Alice's count: 3.0 yr of her own time receding at 1/3 pulse per year gives 1 pulse; 3.0 yr approaching at 3 per year gives 9. Total 10, which is Bob's age at the reunion.
Bob's count: he keeps seeing her recede until light from the turnaround reaches him at D/v + D/c = 5.0 + 4.0 = 9.0 yr, giving 9.0 × (1/3) = 3 pulses; then D/v − D/c = 1.0 yr of approach at 3 per year gives 3 more. Total 6, which is Alice's age.
The asymmetry is now something each twin experiences: Alice's redshift turns blue at the midpoint of her own trip, Bob's only after 90% of his wait. That 9-versus-1 split, not the g-force, is the observational face of the bent worldline.
Generalise: (1/k)(D/v + D/c) + k(D/v − D/c) = (D/v)[(1 + β)/k + k(1 − β)]. Since (1 + β)/k = k(1 − β) = √(1 − β²) = 1/γ, the bracket is 2/γ and the total is 2D/(γv) = T/γ = τ, for any β.

AnswerAlice counts 10 pulses (Bob's 10.0 yr), Bob counts 6 (Alice's 6.0 yr). Both are totals of arrival events, so no frame can dispute them, and the identity gives τ = T/γ for every β.