The Twin Paradox

Two twins. One boards a ship, flies to a star four light-years away at $0.8c$ , turns around, comes home. The other stays on Earth and waits. When the ship lands, the traveller has aged six years and the stay-at-home ten. That is not the paradox. That is just time dilation, and it is measured daily. The paradox is the objection that arrives about four seconds later: motion is relative. From the ship, it is Earth that goes hurtling away at $0.8c$ and comes hurtling back. So the ship's twin should find Earth's clock running slow, and the stay-at-home should be the younger one. Both arguments look identical. Both cannot be right — the twins meet at a single event, stand next to each other, and compare faces. There is exactly one answer. The resolution is not a trick, and it is not about acceleration being "outside" special relativity. It is that the two twins travel along different paths through spacetime, and proper time is the length of a path. They meet again with different ages for the same reason two drivers meeting in the same town can have different odometer readings. But the geometry has a sting in it. In the space you are used to, a straight line is the shortest distance between two points. In spacetime, the straight worldline is the LONGEST — the triangle inequality runs backwards. The twin who does nothing wins, and wins by the maximum possible margin, always.

Key takeaways

Signature $(+,-,-,-)$ , $ds^2 = c^2dt^2 - d\mathbf{x}^2$ , timelike-positive. Proper time is the length of a worldline: $\tau = \int\sqrt{1-\beta(t)^2}\,dt$ (recalled from aq-sr-proper-time).
The situation is NOT symmetric. The stay-at-home is inertial throughout; the traveller is not. Only one of them changes frames, and only one of them feels it. The symmetry argument fails at the turnaround, not at the algebra.
Two-leg answer: $\beta = 0.8$ , $L = 4$ ly $\Rightarrow$ Earth $2L/v = 10$ y, traveller $2L/(v\gamma) = 6$ y. Gap 4 years.
The strictly general theorem (OWNED): for any round trip, with any speed profile, $\tau_{\text{trav}} = \int\sqrt{1-\beta^2}\,dt < T$ strictly. Not "usually". Always.
Owned structural result: the REVERSED triangle inequality. For future-timelike $A, B$ : $A\!\cdot\!B \ge |A||B|$ , hence $\boxed{|A+B| \ge |A| + |B|}$ , with equality iff $A \parallel B$ . Two sides of a Minkowski triangle sum to less than the third. The straight worldline is the longest, and it is the unique maximum.
Key idea: stop asking whose clock runs slow. Ask how long each path is. Time dilation is a statement about rates in a frame; the twins' age gap is a statement about the lengths of two worldlines, and lengths do not need a frame.

The core insight

Proper time is not "time as measured by a moving clock" — it is arc length in spacetime, a property of a path and of nothing else. Two paths between the same two events can have different lengths, and there is no more paradox in that than in two roads between the same two towns having different mileages. Everything that feels strange about the twins is the sign in the metric: the reversed triangle inequality makes the lazy, inertial, do-nothing path the longest one (the maximisation theorem itself — that the straight worldline extremises proper time, and the reversed triangle inequality behind it — is owned by `aq-sr-proper-time-proper-length`; it is used here, not re-proved. What this article owns is the twin scenario: the false symmetry premise, and the numbers).

Where this lives today

Every GPS fix you have ever taken. The satellite clocks run slow by about $7\ \mu\mathrm{s}$ per day from their orbital speed — the twin effect, on a clock that never lands — and fast by about $45\ \mu\mathrm{s}$ per day from the weaker gravity, for a net $+38\ \mu\mathrm{s}$ per day that the system corrects for. Left uncorrected, the positioning error would accumulate at roughly $10$ km per day. The traveller's youth is not a thought experiment; it is a line item in the satellite firmware.

Muons, every second, in the sky above you. A cosmic-ray muon lives $2.2\ \mu\mathrm{s}$ in its own rest frame and should decay long before reaching the ground. It arrives anyway, in numbers, because its worldline is long and its proper time is short. It is the traveller who never comes back, and its arrival at sea level is the twin paradox writing itself on a detector.

And onward into gravity. The statement proved here — the inertial worldline maximises proper time — is the one sentence of special relativity that survives, verbatim, into general relativity. There it becomes the law of motion: a freely falling body follows the worldline of greatest proper time between two events, and gravity is what happens when spacetime is curved enough that the longest path is no longer straight. The twin paradox is the flat-space rehearsal for the geodesic principle, and this is why it is worth doing properly.

The physical picture

The trap is the phrase "motion is relative", used as if it settled the matter. It is true that uniform motion is relative — that is the first postulate. But the twins' situation is not a situation of uniform motion. The traveller's journey has three phases: out, turn, back. The stay-at-home's has one. You can see the asymmetry without any theory at all. Put an accelerometer in each twin's pocket. The stay-at-home's reads zero for the entire experiment. The traveller's reads zero, then spikes hard, then reads zero again. That spike is not a matter of perspective. It is a physical, frame-independent, coffee-spilling event, and it happens to exactly one of the twins. Any argument that treats the twins as interchangeable has already ignored it. So the "symmetric" counter-argument is not subtly wrong. It is a description of a different experiment — one with two inertial observers who separate and never meet again, and in that experiment each really does find the other's clock running slow, with no contradiction, because they never come back to a single event where a comparison could be made. The moment you demand a reunion, you demand that someone turn around, and the symmetry is gone. What replaces it is geometry. Both twins run from event $\mathcal{A}$ (departure) to event $\mathcal{B}$ (reunion). Each one's clock has been ticking off its own path length — the proper time along its own worldline. The question "who is younger" becomes "which path was shorter", and that question has a unique, frame-independent answer, because path length is an invariant.

Interactive Simulation

The Twin Paradox — the bent (accelerated) worldline accumulates less proper time than the straight one, so the traveling twin returns younger.

The core equations

\tau = \int_{\mathcal{A}}^{\mathcal{B}} d\tau = \int_{t_A}^{t_B} \sqrt{1 - \beta(t)^2}\;dt \qquad (1)

where: $\beta(t) = v(t)/c$ is the speed in some inertial frame, and $\tau$ is the time actually elapsed on the clock carried along the path. Physical reading: this is the odometer. It depends on the whole path, not on the endpoints — and the integrand is $\le 1$ , with equality only where $\beta = 0$ . Recalled from aq-sr-proper-time; this article uses it, it does not re-derive it.

t_{\text{Earth}} = \frac{2L}{v}, \qquad \boxed{\;\tau_{\text{trav}} = \frac{2L}{v\gamma} = \frac{t_{\text{Earth}}}{\gamma}\;} \qquad (2)

where: $L$ is the one-way distance in Earth's frame and the trip is two legs at constant speed. Physical reading: the traveller's clock reads less by the factor $1/\gamma$ , and the deficit is real, permanent, and comparable side-by-side at the reunion.

\boxed{\;\tau_{\text{trav}} = \int_0^{T}\sqrt{1-\beta(t)^2}\,dt \;<\; T = \tau_{\text{home}}\;} \qquad\text{for any non-trivial round trip} \qquad (3)

where: $T$ is the elapsed Earth time between departure and reunion. Physical reading: this is the whole theorem, and it is far stronger than the two-leg calculation. No assumption of constant speed, no assumption of a symmetric trip, no assumption about the turnaround being sharp or gentle. Any journey that leaves and returns has $\beta \ne 0$ somewhere, so the integrand dips below 1 somewhere, so the traveller loses. Always.

A\!\cdot\!B \;\ge\; |A|\,|B| \qquad\Longrightarrow\qquad \boxed{\;|A + B| \;\ge\; |A| + |B|\;} \qquad (4)

where: $A$ and $B$ are future-directed timelike four-vectors, $|A| \equiv \sqrt{A\!\cdot\!A}$ , and equality holds iff $A$ and $B$ are parallel. Physical reading: the triangle inequality is reversed. In Euclidean geometry the third side is shorter than the sum of the other two; in Minkowski geometry it is longer. The straight worldline between two timelike-separated events is not the shortest path — it is the longest, and it is the unique longest. That is this article's owned structural result, and it is what the twin paradox is actually about.

Where it all comes from

Derivation AThe trip, and the exact point where the symmetry argument dies

Before any geometry, do the bookkeeping honestly and in one frame. The aim is not just to get the six years — it is to see that the traveller's own accounting agrees, so that "but from the ship's point of view..." cannot be used as an escape hatch. It cannot: the ship's crew get the same answer, for a different-looking reason, and the two reasons are the same reason.

Step 1. Work in Earth's frame, where nothing is subtle. The ship flies out a distance $L$ at speed $v$ , turns, comes back: t_{\text{Earth}} = \frac{2L}{v}. At $\beta = 0.8$ and $L = 4$ ly, that is $8/0.8 = 10$ years. Why: Earth is inertial for the whole experiment, so ordinary kinematics applies with no caveats whatsoever.

Step 2. Get the traveller's clock from equation (1). The speed is $|\beta| = 0.8$ throughout (the sign flips at the turn, but $\beta^2$ does not), so \tau_{\text{trav}} = \int_0^{10}\sqrt{1-0.64}\;dt = 0.6 \times 10 = 6 \text{ years}. Earth 10, traveller 6, gap 4. Why: the integrand is constant here, which is why the two-leg trip is the one every textbook uses. Nothing about the argument requires that.

Step 3. Now do it from the ship, and watch it agree. On the outbound leg the ship is inertial, and in its frame the Earth-to-star distance is Lorentz-contracted: L' = \frac{L}{\gamma} = \frac{4}{5/3} = 2.4 \text{ ly}. The star arrives at the ship at $0.8c$ , so the leg takes $2.4/0.8 = 3$ years of ship time. Same on the way back. Total 6 years — the same 6. Why: the ship's crew do not explain their youth by time dilation at all. They explain it by length contraction: they simply did not go as far. Both frames agree on the reading of the ship's clock, because that reading is an invariant.

Step 4. Locate the asymmetry precisely. Both twins are inertial on the legs. But the traveller is not inertial at the turn: to come home he must change velocity by $2v$ , and his outbound rest frame and his inbound rest frame are different frames. There is no single inertial frame in which the traveller is at rest for the whole trip, and therefore no legitimate way to run the "by symmetry, Earth's clock is the slow one" argument end to end. Why: this is where the paradox dies. The symmetric argument tacitly assumes the traveller has one inertial frame. He has two, and the crime is committed in the act of switching.

Step 5. Refuse the two usual overcorrections. It is not true that special relativity cannot handle acceleration — equation (1) handles any $\beta(t)$ whatsoever, and we just used it. And it is not true that the ageing happens "during" the turnaround: make the turn arbitrarily fast, or arbitrarily gentle, or replace the single traveller with a relay of two inertial ships passing a clock reading between them, with no acceleration anywhere, and the 4-year gap survives unchanged. Why: the gap is a property of the path, not of the kink. The turnaround's only job is to make the path bent. How it does so is irrelevant, which is exactly what equation (3) will say. *Verified in verify_aq-sr-twin-paradox.py: $t_{\text{Earth}} = 10$ y and $\tau_{\text{trav}} = 6$ y at $\beta = 0.8$ , $L = 4$ ly, computed independently in the Earth frame and by the ship's contracted-distance route; $\gamma = 5/3$ ; $L/\gamma = 2.4$ ly; and the per-leg proper time $\sqrt{\Delta t^2 - \Delta x^2} = 3$ y from the invariant interval.*

Go deeper · Derivation B · 2 minProper time is a path length, and the inertial path is the longest

Derivation BProper time is a path length, and the inertial path is the longest

The two-leg result is a special case, and treating it as the whole story is why the paradox keeps coming back to life. Here is the general statement, and it costs one line. It applies to any round trip: constant speed, variable speed, a hundred turnarounds, a leisurely spiral — anything at all, so long as the traveller leaves and returns.

Step 1. Fix the endpoints. Both twins are present at event $\mathcal{A}$ (departure) and at event $\mathcal{B}$ (reunion). The stay-at-home's worldline is the straight one between them; the traveller's is anything else. Let $T$ be the Earth time from $\mathcal{A}$ to $\mathcal{B}$ . Why: the twins must share both endpoints or there is nothing to compare. The paradox is entirely about closed paths.

Step 2. Write each twin's clock as an integral. The stay-at-home has $\beta \equiv 0$ in Earth's frame, so \tau_{\text{home}} = \int_0^T \sqrt{1 - 0}\;dt = T. The traveller has some $\beta(t)$ , so $\tau_{\text{trav}} = \int_0^T \sqrt{1-\beta(t)^2}\;dt$ . Why: both are the same formula, equation (1), applied to two different paths. That is the entire content of the resolution.

Step 3. Compare the integrands, pointwise. For every $t$ , 0 \le \beta(t)^2 < 1 \qquad\Longrightarrow\qquad \sqrt{1-\beta(t)^2} \le 1, with equality iff $\beta(t) = 0$ . Why: no integration is needed to see who wins — the traveller's integrand is never larger than the stay-at-home's, at any instant. The result is already decided before we add anything up.

Step 4. Get the strict inequality. A round trip must leave home and come back, so $\beta(t) \ne 0$ on a set of times of positive duration. On that set the integrand is strictly less than 1. Hence \boxed{\;\tau_{\text{trav}} = \int_0^T\sqrt{1-\beta^2}\,dt \;<\; T = \tau_{\text{home}}.\;} Why: strictly, and for every possible itinerary. There is no clever trajectory, no exotic turnaround, no sequence of accelerations that lets the traveller win or even tie. The stay-at-home's advantage is not typical — it is universal.

Step 5. Name what has been proved. The inertial worldline between two timelike-separated events maximises proper time. It is a geodesic of Minkowski spacetime — and unlike a Euclidean geodesic, which minimises length, this one maximises it. Why: this is the sentence to remember. "Free fall maximises proper time" survives intact into general relativity, where it becomes the equation of motion itself; the twin paradox is the flat-spacetime rehearsal for it. *Verified in verify_aq-sr-twin-paradox.py: $\sqrt{1-\beta^2} \le 1$ with equality iff $\beta = 0$ , symbolically; the strict inequality $\tau_{\text{trav}} < T$ is confirmed by numerical integration for 500 random smooth round-trip speed profiles (including multi-turnaround and variable-speed itineraries), with the maximum found $\tau$ always strictly below $T$ and approaching it only as $\beta \to 0$ .*

Go deeper · Derivation C · 2 minThe reversed triangle inequality

Derivation CThe reversed triangle inequality

Why does the longest path belong to the twin who did nothing? Because the Minkowski metric has a minus sign in it, and that minus sign flips the triangle inequality. This card proves the flip. It is the article's owned structural result, and once you have it, the twin paradox is not a paradox but a corollary — the same corollary as "the straight road has the biggest odometer reading", which is only strange until you notice which geometry you are standing in.

Step 1. Set up the triangle. Let $A$ and $B$ be the two legs of the traveller's trip, as four-vector displacements: $A$ from departure to turnaround, $B$ from turnaround to reunion. Both are future-directed timelike. The stay-at-home's displacement is the third side, $A + B$ . Define the Minkowski length $|A| \equiv \sqrt{A\!\cdot\!A}$ , which is precisely the proper time along that leg. Why: proper time IS length here, so a statement about lengths is a statement about ages.

Step 2. Prove the reversed Cauchy–Schwarz inequality, $A\!\cdot\!B \ge |A||B|$ . Evaluate the invariant $A\!\cdot\!B$ in the rest frame of $A$ , where $A = (|A|, \mathbf{0})$ and $B = (b^0, \mathbf{b})$ . Timelike and future-directed means $b^0 = \sqrt{|B|^2 + |\mathbf{b}|^2} \ge |B|$ . Therefore A\!\cdot\!B = |A|\,b^0 \;\ge\; |A|\,|B|, with equality iff $\mathbf{b} = \mathbf{0}$ , i.e. iff $B$ is parallel to $A$ . Why: note the direction of the inequality — in Euclidean space Cauchy–Schwarz gives $\mathbf{a}\cdot\mathbf{b} \le |\mathbf{a}||\mathbf{b}|$ . The minus signs in the metric turn the bound upside down, and everything downstream follows from that one reversal.

Step 3. Square the third side. |A+B|^2 = (A+B)\!\cdot\!(A+B) = |A|^2 + 2\,A\!\cdot\!B + |B|^2 \;\ge\; |A|^2 + 2|A||B| + |B|^2 = \big(|A| + |B|\big)^2, using Step 2. Why: pure algebra — the physics was all spent in Step 2.

Step 4. Take the square root. $A + B$ is future-timelike (a sum of future-timelike vectors is future-timelike), so both sides are positive and the root is safe: \boxed{\;|A + B| \;\ge\; |A| + |B|, \qquad \text{equality iff } A \parallel B.\;} Why: this is the triangle inequality with the sign reversed. Two sides of a Minkowski triangle sum to LESS than the third. The stay-at-home's single straight side beats the traveller's two, and the equality case tells you exactly when the twins tie — never, unless the traveller never turns, in which case there was no trip.

Step 5. Read the twin paradox off the geometry. $|A+B| = \tau_{\text{home}}$ and $|A| + |B| = \tau_{\text{trav}}$ . The inequality says $\tau_{\text{home}} \ge \tau_{\text{trav}}$ , strictly whenever the legs are not parallel — i.e. whenever the traveller actually turned around. At $\beta = 0.8$ , $L = 4$ ly: $|A| = |B| = 3$ y, $|A|+|B| = 6$ y, while $|A+B| = 10$ y. $6 < 10$ : two sides sum to less than the third. Why: iterate the inequality over a polygon and you recover equation (3) for any piecewise-inertial path, then take the continuum limit for any path at all. The reversed triangle inequality is the twin paradox, stripped of its story. *Verified in verify_aq-sr-twin-paradox.py: the reversed Cauchy–Schwarz $A\!\cdot\!B \ge |A||B|$ and the reversed triangle inequality $|A+B| \ge |A|+|B|$ are proved symbolically in the rest frame of $A$ and then confirmed on 200,000 random pairs of future-timelike four-vectors (zero violations); the equality case is confirmed to hold exactly for parallel vectors and to fail otherwise; and the $\beta = 0.8$ triangle gives $3 + 3 = 6 < 10$ .*

Figure

PUNCHLINE — the twin paradox is a triangle. Earth's worldline is the straight vertical side, 10 years long. The traveller's two legs are 3 years each, 6 in total. Every length here is a PROPER TIME, computed as √(Δt² − Δx²), and it is what the clock on that path actually reads. Two sides of a Minkowski triangle sum to LESS than the third: 3 + 3 = 6 < 10. In Euclidean geometry the straight line is the SHORTEST path between two points. In spacetime it is the LONGEST — which is why the twin who did nothing at all comes back the oldest

Figure

The theorem is not about the two-leg trip. For a round trip of FIXED Earth duration T = 10 years, the traveller's clock reads τ = T√(1 − β²) — and that is below 10 for EVERY β greater than zero. There is no itinerary that lets the traveller win, or even tie. The curve is the proper-time deficit as a function of how hard the traveller flies: β = 0.6 → 8 y; β = 0.8 → 6 y; β = 0.95 → 3.1225 y; β = 0.99 → 1.4107 y. As β → 1 the traveller's clock stops altogether and he returns having aged nothing, to an Earth ten years older

Figure

THE OWNED RESULT, in one picture. LEFT — Euclidean: the straight side is the SHORTEST, 3 + 4 = 7 > 5, and the triangle inequality reads |A + B| ≤ |A| + |B|. RIGHT — Minkowski: the straight side is the LONGEST, 3 + 3 = 6 < 10, and the SAME inequality runs BACKWARDS, |A + B| ≥ |A| + |B|. Nothing changed but a minus sign in the metric. Every strange thing about the twins is downstream of that one sign

Limiting cases

Case	Result	Reading
$\beta \to 0$	$\tau_{\text{trav}} \to T$	no trip, no gap; the equality case of (4)
$\beta \to 1$	$\tau_{\text{trav}} \to 0$	the traveller returns having aged nothing
$\beta \ll 1$	$\tau_{\text{trav}} \approx T(1 - \tfrac12\beta^2)$	the gap is second order — why nobody noticed for 250 years
any $\beta(t) \not\equiv 0$	$\tau_{\text{trav}} < T$ , strictly	equation (3): no itinerary escapes
legs parallel	$\|A+B\| = \|A\|+\|B\|$	equality in (4): the traveller never turned
$\beta = 0.8$ , $L = 4$ ly	$10$ y vs $6$ y	the canonical numbers

Common trapThe turnaround jump is 2Lv/c² — with NO gamma

Base defect D5. If you resolve the paradox by the traveller's changing lines of simultaneity — which is the method owned by aq-sr-twin-paradox-asymmetry, not this article — the simultaneity gap at the turn is \Delta t_{\text{jump}} = \frac{2Lv}{c^2}, with no factor of $\gamma$ . The AllQFT base prints a spurious $\gamma$ here, and the error is trivially caught by making the books close. The traveller reckons Earth's clock to advance $L/(v\gamma^2)$ per leg, so \frac{2L}{v\gamma^2} + \frac{2Lv}{c^2} = \frac{2L}{v} \quad\checkmark At $\beta = 0.8$ , $L = 4$ ly: $3.6 + 6.4 = 10$ years, which is exactly Earth's elapsed time. With the base's spurious $\gamma$ the jump becomes $10.667$ y and the total $14.27$ y — more than the trip took. (A companion defect: the base's own equation card also gives the per-leg Earth time as $(L/v)\gamma^{-1}$ ; it is $(L/v)\gamma^{-2}$ , as its own derivation says two lines later.) Always make the traveller's books close against $2L/v$ .

Common trapAcceleration is not the answer, it is only the excuse

It is very common to hear that the paradox is resolved "because the traveller accelerates, and acceleration is outside special relativity". Both halves of that are wrong. Special relativity handles acceleration perfectly well — equation (1) takes any $\beta(t)$ — and the age gap is not produced by the acceleration. Make the turnaround take a microsecond or a decade, make it a gentle arc or a violent kick: the gap is unchanged, because it is fixed by the shape of the path, not by what happened at the corner. The cleanest demonstration removes acceleration entirely: let an outbound ship pass its clock reading to an inbound ship at the turnaround point. Nothing accelerates, nobody feels a thing, and the 4-year gap is still there. Acceleration's only role is to be the thing that makes the path bent — it is the reason the twins are not symmetric, not the mechanism of the ageing.

Worked examples

1 · Four light-years at 0.8c, and the real Proxima Centauri

A ship flies to a star $L = 4$ ly away at $\beta = 0.8$ , turns instantly, and returns. Then repeat for the nearest real star.

Step 1 — identify. Two-leg trip at constant speed: equation (2). $\gamma = 1/\sqrt{1-0.64} = 1/0.6 = 5/3$ .

Step 2 — Earth's clock. $t_{\text{Earth}} = 2L/v = 8/0.8 = 10$ years.

Step 3 — the ship's clock. $\tau = t_{\text{Earth}}/\gamma = 10 \times 0.6 = 6$ years. The traveller returns 4 years younger than the twin. Cross-check from the invariant interval, per leg: $\sqrt{(5\ \text{y})^2 - (4\ \text{ly})^2} = \sqrt{25-16} = 3$ years, and $3 + 3 = 6$ ✓.

Step 4 — the real star. Proxima Centauri lies $4.2465$ ly away. The same trip gives $t_{\text{Earth}} = 2(4.2465)/0.8 = 10.6163$ years and $\tau = 0.6 \times 10.6163 = 6.3698$ years — a gap of $4.2465$ years. (The gap happens to equal $L$ in these units, because $(2/\beta)(1 - 1/\gamma) = 1$ exactly at $\beta = 0.8$ .)

Step 5 — the boundary number. A four-year age difference between siblings, from a decade's round trip to the nearest star, at a speed no engine can approach. The effect is enormous in principle and invisible in practice, and the gap between those two facts is the entire history of people refusing to believe it.

2 · An asymmetric trip, and the reversed triangle inequality with numbers

Now break the symmetry of the legs. The traveller goes out at $\beta_1 = 0.6$ and returns at $\beta_2 = 0.95$ , arriving back at Earth exactly $T = 10$ years after departure. How much has he aged?

Step 1 — identify. Two inertial legs of different speeds. Equal distances out and back: $\beta_1 t_1 = \beta_2 t_2$ , with $t_1 + t_2 = 10$ .

Step 2 — solve for the legs. $t_2 = (0.6/0.95)\,t_1 = 0.631579\,t_1$ , so $t_1(1.631579) = 10$ : t_1 = 6.1290 \text{ y}, \qquad t_2 = 3.8710 \text{ y}. The turnaround happens at $x = 0.6 \times 6.1290 = 3.6774$ ly (and $0.95 \times 3.8710 = 3.6774$ ly ✓ — the legs meet, as they must).

Step 3 — apply equation (1), leg by leg. $\gamma_1 = 1.25$ so $1/\gamma_1 = 0.8$ ; $\gamma_2 = 3.2026$ so $1/\gamma_2 = 0.3122$ . \tau = \frac{t_1}{\gamma_1} + \frac{t_2}{\gamma_2} = 6.1290(0.8) + 3.8710(0.3122) = 4.9032 + 1.2087 = \boxed{6.1119 \text{ years}.}

Step 4 — check the inequality. $|A| = 4.9032$ y, $|B| = 1.2087$ y, $|A| + |B| = 6.1119$ y. And $|A + B| = 10$ y, the straight side. Indeed |A+B| = 10 \;\ge\; |A| + |B| = 6.1119 \quad\checkmark exactly as equation (4) demands — and strictly, because the legs are not parallel.

Step 5 — the boundary number. Note what did not happen: nothing about the trip was symmetric, the legs had different speeds and different durations, and the answer still came out below 10 years. That is the point of the general theorem. The traveller's loss does not depend on the shape of the trip, only on the fact that there was one.

Physics of the portrait

Origins · 2 minTwin paradox and how this was discovered

Twin paradox

1905–1971 · Bern · Paris · Berlin · Washington

The twin paradox was never posed by the man who created it. In §4 of Zur Elektrodynamik bewegter Körper (1905), Einstein noted, almost in passing, that a clock carried around a closed curve and returned to its starting point will lag behind one that stayed at rest — a remark of two sentences, presented as a curiosity, with no twins and no paradox. It was Paul Langevin who supplied the human being. In a lecture at Bologna on 10 April 1911, published in Scientia, he described a traveller who flies away at a speed within one part in twenty thousand of $c$ , turns, and comes home: two years pass for him, and two hundred for the Earth he left. Langevin used the story to argue that the effect was real, not an artefact of bookkeeping, and he located the asymmetry correctly — in the traveller's change of motion. Max von Laue, in 1913, gave the argument the form used here: the traveller's worldline is broken, the stay-at-home's is straight, and Minkowski's geometry decides the rest. The word "paradox" and the twins themselves are later accretions, attached to the problem by decades of people who found the conclusion intolerable. It stopped being arguable in 1971, when Hafele and Keating flew caesium clocks around the world in both directions and brought them home disagreeing with the ground — by nanoseconds, and in exactly the predicted amounts. The traveller really does come back younger. He always did.

History

1905

— Einstein notes in section 4 of Zur Elektrodynamik bewegter Körper that a clock carried around a closed curve lags behind one left at rest. He does not call it a paradox and there are no twins.

1911

— Langevin, in a lecture at Bologna on 10 April published in Scientia, gives the traveller a human form: two years for him, two hundred for the Earth, at a speed one part in twenty thousand below c.

1911

— von Laue, in December, resolves the asymmetry geometrically in Minkowski spacetime: among all timelike worldlines joining two events the straight one has the greatest proper time. In 1913 he adds the frame-switching reading — the traveller uses two inertial frames, the stayer one — locating the asymmetry there rather than in acceleration per se.

1971

— Hafele and Keating fly caesium clocks around the world and back. Eastward: minus 59 plus or minus 10 nanoseconds observed against minus 40 plus or minus 23 predicted. Westward: plus 273 plus or minus 7 against plus 275 plus or minus 21.

2010

— Chou, Hume, Rosenband and Wineland see time dilation with optical clocks at a walking pace of 10 metres per second, and across a height difference of 33 centimetres.

Going further

Texts. Taylor & Wheeler, Spacetime Physics, 2nd ed., ch. 4 (the definitive treatment of the reversed triangle inequality) · Rindler, Relativity: Special, General and Cosmological, 2nd ed., §3.4 · Misner, Thorne & Wheeler, Gravitation, §1.3 · Langevin, "L'évolution de l'espace et du temps", Scientia 10, 31 (1911) · Hafele & Keating, Science 177, 166 and 168 (1972).

Provenance. Signature $(+,-,-,-)$ , re-derived. All numbers recomputed in verify_aq-sr-twin-paradox.py, including the asymmetric trip and the 200,000-sample test of the reversed triangle inequality. Base defect repaired. D5: the base prints the turnaround simultaneity jump with a spurious factor of $\gamma$ ; it is $2Lv/c^2$ , and the books close only without the $\gamma$ ( $3.6 + 6.4 = 10$ y at $\beta = 0.8$ , $L = 4$ ly). A companion error in the base's own equation card gives the per-leg Earth time reckoned by the traveller as $(L/v)\gamma^{-1}$ instead of $(L/v)\gamma^{-2}$ — contradicting its own derivation two lines later.

Scope. This article owns the setup and the proper-time (geometric) resolution. It does not own the other two standard resolutions, and does not re-derive them: the simultaneity-jump account belongs to aq-sr-twin-paradox-asymmetry, and the Doppler pulse-counting account to aq-sr-twin-paradox-doppler-resolution. Both are recalled here by name only.

Builds on. Time Dilation · The Relativity of Simultaneity · The Lorentz Transformation Equations · Proper Time. Leads to. The Asymmetry in the Twin Paradox · The Doppler-Shift Resolution of the Twin Paradox · Longitudinal Relativistic Doppler Effect · Rapidity and Additive Velocity.

If motion is relative, why is the traveller not equally entitled to say that Earth moved?

He is entitled to say it — on each leg separately, and while he says it he is right. What he is not entitled to do is say it for the whole trip, because there is no single inertial frame in which he is at rest from departure to reunion. He has one rest frame going out and a different one coming back. The stay-at-home has one, throughout. The "symmetric" argument quietly assumes the traveller has a single frame, and that assumption is false — which is why the argument produces a contradiction. Motion is relative; changing your motion is not.

Is the resolution that acceleration is outside special relativity?

No, and this is the most persistent piece of misinformation about the problem. Special relativity handles accelerated motion perfectly well: equation (1) accepts any $\beta(t)$ , and we integrated it. Worse, the "acceleration did it" story makes a prediction that is simply wrong — it suggests that a harder or longer turnaround should change the age gap. It does not. Keep the endpoints and the legs fixed and vary the turnaround however you like: the gap is unchanged, because the gap is the length of the path, and the corner is a set of measure zero.

Does the traveller see Earth's clock running slow? If so, how does Earth end up ahead?

Yes, on each leg, and yes, it still ends up ahead. On the outbound leg the traveller genuinely reckons Earth's clock to be ticking slowly, and the same on the way back. But "reckons" involves a line of simultaneity, and the traveller's line swings through a large angle at the turnaround, sweeping across a great swathe of Earth's history that belonged to neither leg. Add the two legs' reckoning to the swing and you recover Earth's full elapsed time exactly. That accounting is the subject of aq-sr-twin-paradox-asymmetry; the warning above gives the number and the books-closing check, since the base gets it wrong.

Is the age difference real, or just an artefact of comparing clocks in different frames?

Real, permanent, and comparable side by side. The two twins are together at a single event when the ship lands — same place, same time, one frame, no coordinate subtlety of any kind. Their clocks read different numbers, their faces are different ages, and there is nothing left to be relative about. This is precisely why the reunion matters: it converts a statement about frames into a statement about two numbers on two clocks sitting on the same table.

Has anyone actually measured this?

Yes. In 1971 Hafele and Keating flew caesium clocks around the world eastward and westward and compared them with clocks that stayed at the observatory. The travelling clocks came home disagreeing with the stay-at-home clocks — the eastward flight lost $59 \pm 10$ ns against a prediction of $40 \pm 23$ ns, the westward gained $273 \pm 7$ ns against $275 \pm 21$ ns. In 2010 Chou and co-workers saw the effect with optical clocks at a walking pace of $10\ \mathrm{m\,s^{-1}}$ . And every GPS satellite is a twin-paradox experiment running continuously, whose result is corrected for in firmware.

Why is the longest path the lazy one? That is the opposite of every other geometry I know.

Because of the minus sign. In the Euclidean plane, adding a detour adds a positive contribution to $\Delta x^2 + \Delta y^2$ , so wandering makes a path longer. In spacetime, the interval is $c^2\Delta t^2 - \Delta x^2$ , so adding spatial wandering subtracts, and makes the path shorter. Motion costs proper time rather than earning it. Run that reasoning through Cauchy–Schwarz and it comes out as the reversed triangle inequality $|A+B| \ge |A|+|B|$ — the straight worldline is the unique maximum. Everything strange about the twins is downstream of that one sign in the metric.

The Twin Paradox

Where this lives today

The physical picture

The core equations

Where it all comes from

Limiting cases

Worked examples

1 · Four light-years at 0.8c, and the real Proxima Centauri

2 · An asymmetric trip, and the reversed triangle inequality with numbers

Physics of the portrait

Going further

Before this

Where this leads