Newton's Third Law

A rocket engine firing in deep space has, by everyday intuition, nothing to push against — no air, no road, no launch pad. Yet it accelerates, and the reason was written down in 1687: forces never come singly. Whenever one body pushes or pulls on another, the second pushes or pulls back on the first with exactly equal magnitude and exactly opposite direction. The rocket hurls exhaust gas backward; the gas, in the very act of being hurled, pushes the rocket forward. Nothing external is needed, because the pair is complete in itself. Newton's Third Law is the shortest of the three laws and the most widely misquoted. The trap hides in plain sight: the two forces of a pair act on different bodies. The cart really does pull back on the horse exactly as hard as the horse pulls the cart — and they move anyway, because the backward pull acts on the horse while the forward pull acts on the cart, and no free-body diagram ever contains both. Learn to place each force on the right body and a whole family of "paradoxes" dissolves at once. There is a deeper payoff. Sum Newton's Second Law over two interacting bodies and the Third Law makes the internal forces cancel: total momentum stops changing. The Third Law is momentum conservation, read at a single instant — which is why it survives, suitably generalized, into electromagnetism, relativity, and quantum field theory, long after "force" itself has faded from the fundamental vocabulary.

Key takeaways

If body A exerts $\mathbf{F}_{AB}$ on body B, then B exerts $\mathbf{F}_{BA}=-\mathbf{F}_{AB}$ on A — equal magnitude, opposite direction, at every instant.
The two forces of a pair act on different bodies: they never appear in the same free-body diagram and never cancel each other.
Adding the Second-Law equations of an isolated pair makes the internal forces cancel: total momentum $m_A\mathbf{v}_A + m_B\mathbf{v}_B$ is conserved.
The strong form adds collinearity — pair forces point along the line joining the bodies — and that extra property is what conserves total angular momentum.
Magnetic forces between moving charges violate the law; conservation survives because the electromagnetic field itself carries momentum.
Key idea: forces come in equal-and-opposite pairs acting on different bodies; this pairing is mathematically equivalent to conservation of total linear momentum for an isolated system.

Where this lives today

Every spacecraft is a Third-Law machine. Chemical rockets, the ion engines that propelled Dawn to the asteroid belt on millinewtons of thrust, the Hall thrusters steering thousands of Starlink satellites, the cold-gas jets astronauts once wore — all of them accelerate by throwing mass one way and pocketing the reaction. No medium is needed and none is used; the exhaust itself is the thing pushed against. Reaction wheels run the angular version of the same trade: spin a flywheel one way and the telescope turns the other, total angular momentum fixed at zero — how Hubble and JWST point without burning a gram of fuel. On the ground, the law is measurement infrastructure. An atomic force microscope reads the reaction on its cantilever to map single atoms; force plates in gait labs record the ground's push on runners; strain gauges in tow couplings and robot joints read the single interface force whose two addresses you computed above. And the "failure" of the law became physics' gain: the momentum missing from magnetic force pairs sits in the electromagnetic field, arriving as real radiation pressure — the push that optical tweezers (2018 Nobel Prize) use to hold living cells, and that laser-cooling rigs use to freeze atoms to microkelvins.

The physical picture

Push on a wall and you feel the wall push back on your hand — harder as you press harder, vanishing the instant you let go. You cannot touch anything without being touched by it, with exactly the same strength. That mutual, simultaneous grip is the Third Law made tangible, and it holds for every kind of force: contact, tension, friction, gravity. Walking runs on it. Your foot pushes backward on the ground; the ground pushes forward on you, and that forward reaction is the only horizontal external force available to accelerate your body. You move not because of the force you exert — that one acts on the planet — but because of the force the planet returns. On frictionless ice the trick fails: you can press straight down all you like, but the reaction has no horizontal component to grip, so you go nowhere. The whole art of using the law is bookkeeping: every force belongs to some body, and the reaction always belongs to the other body. The classic sorting exercise is a book resting on a table — three bodies (book, table, Earth), two genuine pairs, and one impostor:

Figure

One balance and two true pairs — sorting the forces on a book at rest

Read the figure left to right. The left panel shows the two forces on the book: the normal force up, gravity down. They balance — but only because the book happens to be in equilibrium, and they are not even the same kind of force. The middle and right panels show the two genuine Third-Law pairs: the contact pair (table pushes book up, book pushes table down) and the gravity pair (Earth pulls book down, book pulls Earth up). Each pair straddles two bodies; no pair ever lives inside one free-body diagram. The same bookkeeping, run for two ice skaters pushing apart, delivers the law's most striking experimental signature — recoil in inverse proportion to mass:

Figure

Recoil — equal and opposite forces, equal and opposite momenta

Whatever momentum one skater gains, the other gains its negative, because the forces they exchange are equal, opposite, and last exactly as long as the contact does. The lighter skater leaves faster — speeds in inverse proportion to the masses — but the momentum ledger stays perfectly balanced. Keep that ledger in mind: everything in the mathematics below is that picture, formalized.

Pairs, not balances

Everything begins with a naming convention worth internalizing: write $\mathbf{F}_{AB}$ for "the force that A exerts on B." The first index is the agent, the second the receiver. With the notation fixed, the entire content of the Third Law fits on one line. Newton phrased it as Lex III: to every action there is always opposed an equal reaction; in modern vector form,

\mathbf{F}_{BA}=-\mathbf{F}_{AB} \qquad (1)

where: $\mathbf{F}_{AB}$ — force that body A exerts on body B; $\mathbf{F}_{BA}$ — force that body B exerts on body A. Taking magnitudes gives $|\mathbf{F}_{AB}|=|\mathbf{F}_{BA}|$ automatically: neither body is "stronger." Note what the equation does not say — it never adds the two forces, because they act on different bodies and belong to different free-body diagrams. This is the weak form of the law, and it holds for contact forces, tension, friction, springs, and Newtonian gravity alike. Many interactions obey something stronger. For gravity and electrostatics the pair forces do not merely oppose each other — they point along the line joining the two bodies. That extra geometric property gets its own name, the strong form:

\mathbf{F}_{AB}=f(r)\,\hat{\mathbf{r}}_{AB},\qquad \hat{\mathbf{r}}_{AB}=\frac{\mathbf{r}_B-\mathbf{r}_A}{|\mathbf{r}_B-\mathbf{r}_A|} \qquad (2)

where: $f(r)$ — scalar function of the separation $r$ (positive for repulsion); $\hat{\mathbf{r}}_{AB}$ — unit vector from A toward B; $\mathbf{r}_A,\mathbf{r}_B$ — position vectors. The strong form looks like a technicality, but it is the exact ingredient that conserves angular momentum, as we will prove below. Magnetic forces between moving charges satisfy neither form — the first crack in the Newtonian edifice, and a signpost toward field theory.

From pairs to conserved momentum

Why does this one sign convention deserve the status of a law of nature? Because it is secretly a conservation law. Take two bodies interacting only with each other — an isolated pair, no external forces — and watch what the Third Law does to their combined momentum.

Derivation AFrom the Third Law to conserved total momentum

The target is equation (3): for two bodies interacting only with each other, the total momentum $m_A\mathbf{v}_A+m_B\mathbf{v}_B$ never changes, no matter how violent or complicated the interaction. The route is the shortest one in mechanics — write the Second Law for each body separately, add the two equations, and let the Third Law kill the right-hand side. We choose this route because it exposes exactly which ingredient does the work: addition assembles the total momentum, but only the pairing $\mathbf{F}_{AB}=-\mathbf{F}_{BA}$ makes its derivative vanish. Watch the scope of the cancellation — each body's individual momentum changes throughout the interaction; it is only the sum that the Third Law protects. Keep track of which force acts on which body, because swapping the labels $\mathbf{F}_{AB}$ and $\mathbf{F}_{BA}$ is the classic way this three-line argument goes wrong.

Step 1. Write the Second Law for each body. $m_A\dot{\mathbf{v}}_A=\mathbf{F}_{BA}$ and $m_B\dot{\mathbf{v}}_B=\mathbf{F}_{AB}$ . Why: each body feels only the force from its partner, so its momentum changes at exactly the rate that force dictates.

Step 2. Add the two equations. The left side becomes $\frac{\mathrm{d}}{\mathrm{d}t}(m_A\mathbf{v}_A+m_B\mathbf{v}_B)=\frac{\mathrm{d}\mathbf{p}_{\text{tot}}}{\mathrm{d}t}$ , the right side $\mathbf{F}_{BA}+\mathbf{F}_{AB}$ . Why: addition is legal — these are two vector equations holding simultaneously — and it assembles precisely the quantity whose fate we care about, the total momentum.

Step 3. Invoke the Third Law. By (1) the right side is $\mathbf{F}_{BA}+\mathbf{F}_{AB}=0$ , so

\mathbf{p}_{\text{tot}}=m_A\mathbf{v}_A+m_B\mathbf{v}_B=\text{const} \qquad (3)

Why: the two internal forces are equal and opposite, so their sum — and only their sum — vanishes. Individually each body's momentum changes constantly during the interaction; it is the total that the pairing protects.

where: $m_A,m_B$ — the two masses; $\mathbf{v}_A,\mathbf{v}_B$ — their velocities; $\mathbf{p}_{\text{tot}}$ — total linear momentum. This three-line argument is the entire logical content of the Third Law: pairing of forces = conservation of momentum. Run it backward and conservation implies the pairing — the two statements are equivalent. The argument scales up without modification. In a system of many particles, every internal force belongs to some pair $(i,j)$ , and each pair cancels internally:

\sum_{i}\mathbf{F}^{\text{int}}_{i}=\sum_{i<j}\left(\mathbf{F}_{ij}+\mathbf{F}_{ji}\right)=0 \qquad (4)

where: $\mathbf{F}_{ij}$ — force on particle $i$ due to particle $j$ ; the sum runs over distinct interacting pairs. However complicated the internal wiring — springs, collisions, gravity between every pair — the internal forces contribute exactly nothing to the total. That cancellation is what makes the centre of mass such a powerful concept. Sum the Second Law over all particles: the internal forces drop out by (4), and what survives is a single-particle equation for the system as a whole,

M\ddot{\mathbf{R}}_{\text{cm}}=\mathbf{F}_{\text{ext}} \qquad (5)

where: $M=\sum_i m_i$ — total mass; $\mathbf{R}_{\text{cm}}$ — centre-of-mass position; $\mathbf{F}_{\text{ext}}$ — net external force. A wrench tumbling through the air is a mess of internal stresses, yet its centre of mass traces a clean parabola — the Third Law has silenced every internal force. This theorem is why "treat the object as a point particle" ever works.

Angular momentum needs the strong form

Linear momentum needed only the weak form. Angular momentum is more demanding: internal torques must cancel too, and equal-and-opposite is not enough — a pair of equal and opposite forces that are offset sideways form a couple, and a couple has net torque. Here is exactly where collinearity earns its keep.

Go deeper · Derivation B · 1 minWhy internal torques cancel only for collinear pairs

Derivation BWhy internal torques cancel only for collinear pairs

We are after equation (6): the total angular momentum of a system changes only through external torque, $\mathrm{d}\mathbf{L}_{\text{tot}}/\mathrm{d}t=\boldsymbol{\tau}_{\text{ext}}$ — and, more pointedly, after the reason this demands the strong form of the Third Law rather than the weak one. The strategy is to write the total torque about a fixed origin, isolate the contribution of a single interacting pair, and see exactly what it takes to make that contribution vanish. The weak form gets us halfway: it lets the force factor out, collapsing the pair's torque to $(\mathbf{r}_A-\mathbf{r}_B)\times\mathbf{F}_{BA}$ . The watch-out is precisely here — equal-and-opposite forces that are offset sideways form a couple with net torque, so this cross product does not vanish for free; only collinearity of the force with the line of centres finishes the job. If you have ever wondered why textbooks bother stating two forms of the Third Law, this derivation is the answer.

Step 1. Write the total torque. $\frac{\mathrm{d}\mathbf{L}}{\mathrm{d}t}=\sum_i \mathbf{r}_i\times \mathbf{F}_i$ , summing $\mathbf{r}\times\mathbf{F}$ over all particles about a fixed origin. Why: this is the rotational analogue of the Second Law, before any assumption about the forces.

Step 2. Isolate one pair's contribution. $\mathbf{r}_A\times\mathbf{F}_{BA}+\mathbf{r}_B\times\mathbf{F}_{AB}=(\mathbf{r}_A-\mathbf{r}_B)\times\mathbf{F}_{BA}$ . Why: the weak form lets us write $\mathbf{F}_{AB}=-\mathbf{F}_{BA}$ and factor the force out of both terms — the pair's torque depends only on the separation vector, not on where the origin sits.

Step 3. Demand collinearity. If $(\mathbf{r}_A-\mathbf{r}_B)\parallel \mathbf{F}_{BA}$ — the strong form — the cross product vanishes: $(\mathbf{r}_A-\mathbf{r}_B)\times\mathbf{F}_{BA}=0$ . Why: a cross product of parallel vectors is zero; if instead the pair forces were offset from the line of centres, this term would survive as an internal couple and internal forces could spin the system up from nothing. Every internal pair thus contributes zero torque, and only the outside world can change the system's angular momentum:

\frac{\mathrm{d}\mathbf{L}_{\text{tot}}}{\mathrm{d}t}=\boldsymbol{\tau}_{\text{ext}} \qquad (6)

where: $\mathbf{L}_{\text{tot}}$ — total angular momentum about a fixed point; $\boldsymbol{\tau}_{\text{ext}}$ — net external torque. This is why an isolated spinning system — a figure skater, a galaxy, a planet-moon pair — keeps its total angular momentum no matter how violently its parts interact.

The impulse view — collisions in one line

Collisions are where the Third Law does its quietest, most reliable work. During a crash the contact force is enormous, messy, and unknown as a function of time — and none of that matters. Integrate (1) over the duration of contact:

\int_0^{\Delta t}\mathbf{F}_{AB}\,\mathrm{d}t=-\int_0^{\Delta t}\mathbf{F}_{BA}\,\mathrm{d}t \qquad (7)

where: $\Delta t$ — duration of the interaction; each integral is the impulse delivered to one body. Whatever the force profile — a spike, a plateau, a rattle — the two bodies receive equal and opposite impulses, hence equal and opposite momentum changes. Every collision formula you will ever use descends from this line; it is why momentum conservation holds through a car crash even though nobody can model the crumpling.

Chains of pairs — the stacked blocks

When several bodies touch, each interface carries its own Third-Law pair, and the pairs chain together like links. The cleanest laboratory is two stacked blocks: block 1 (mass $m_1$ ) rests on block 2 (mass $m_2$ ), which rests on the ground. Write $N_{12}$ for the force block 1 exerts on block 2, and trace every force to its body:

Figure

Three free-body diagrams, two color-coded pairs at the interfaces

Go deeper · Derivation C · 2 minSolving the stack, one interface at a time

Derivation CSolving the stack, one interface at a time

The goal is every contact force in the two-block stack — in particular the force the ground supplies, which we will show is $(m_1+m_2)g$ , the full combined weight, even though the ground touches only block 2. The method is a strict alternation: apply equilibrium (the Second Law with zero acceleration) within one body's free-body diagram, then use the Third Law to cross an interface to the neighbouring body, and repeat down the chain. We work this way because it is the only bookkeeping that never cheats — each equation involves forces on exactly one body, and the Third Law is invoked only to hop between diagrams. The trap to avoid is mixing the two moves: writing a force on block 1 and a force on block 2 in the same balance, or assuming the ground "feels block 1 directly" rather than through the single interface it actually touches. Done honestly, each new contact adds exactly one Third-Law pair, and the unknowns fall one per step.

Step 1. Free-body diagram of block 1. $N_{21}-m_1 g=0\;\Rightarrow\;N_{21}=m_1 g$ . Why: block 1 touches only block 2 and feels only gravity besides, and it is in equilibrium, so the upward push from below must balance its own weight — nothing else can.

Step 2. Cross the 1–2 interface with the Third Law. $N_{12}=N_{21}=m_1 g$ . Why: "block 2 pushes block 1 up" and "block 1 pushes block 2 down" are one interaction seen from its two ends — equation (1) forces their magnitudes to match.

Step 3. Free-body diagram of block 2. $N_{2g}$ up, gravity $m_2 g$ down, and the load $N_{12}$ down: $N_{2g}-m_2 g - N_{12}=0$ . Why: block 2 has two neighbours, so its diagram carries two contact forces — the ground's support and block 1's press — plus its own weight.

Step 4. Solve for the ground force. $N_{2g}=m_2 g + m_1 g=(m_1+m_2)g$ . Why: block 2 must hold up not just itself but everything stacked on it; the loads accumulate downward through the chain of interfaces.

Step 5. Cross the 2–ground interface. $N_{g2}=N_{2g}=(m_1+m_2)g$ : block 2 presses on the ground with the full combined weight, and the ground presses back equally. Why: one more application of (1); the ground never "knows" about block 1 directly — it feels the whole stack only through the single interface it touches. The same interface-by-interface logic is why a massless rope transmits one tension throughout its length, and why every contact you add to a statics problem brings exactly one new pair — halving the unknowns you might have feared.

Limiting cases

Case	Result	Meaning
Contact and gravity (weak form)	$\mathbf{F}_{BA}=-\mathbf{F}_{AB}$	Equal and opposite; total linear momentum of the isolated pair is conserved. Holds for normal force, tension, friction, springs, Newtonian gravity.
Central forces (strong form)	$\mathbf{F}_{AB}=f(r)\,\hat{\mathbf{r}}_{AB}$	Additionally collinear with the line of centres — the ingredient that cancels internal torques and conserves total angular momentum.
Magnetic forces between moving charges	$\mathbf{F}_{AB}+\mathbf{F}_{BA}\neq 0$	Velocity-dependent Lorentz forces violate even the weak form; conservation survives only by crediting the field with momentum $\int \varepsilon_0\,\mathbf{E}\times\mathbf{B}\,\mathrm{d}V$ .
Retarded interactions	$\mathbf{F}_{AB}(t)$ depends on B at $t-r/c$	With finite signal speed, action and reaction are not even simultaneous; the instantaneous Third Law is the $v\ll c$ , negligible-retardation approximation.

Common trap“Action and reaction cancel, so nothing can ever accelerate”

The two forces of a pair act on two different objects, so they never appear together in one free-body diagram and never cancel each other. To apply $\mathbf{F}=m\mathbf{a}$ to a body you sum only the forces on that body; the reaction lives in the other body's diagram. The forces that do balance on a body in equilibrium — gravity and the normal force on a resting book — are a Second-Law balance, NOT a Third-Law pair: they act on the same body, they are different kinds of force, and their equality breaks the moment the book accelerates (in a dropping elevator, $N < mg$ instantly, while each true pair stays perfectly matched). The genuine reactions are the book's gravitational pull on the Earth and the book's push on the table.

Worked examples

1 · Two skaters push apart (confidence builder)

Skaters of mass $48\ \mathrm{kg}$ and $72\ \mathrm{kg}$ stand at rest on smooth ice, palms touching, then push apart. The heavier skater glides away at $0.90\ \mathrm{m/s}$ . Find the lighter skater's speed, each skater's momentum, and the average force if the push lasted $0.60\ \mathrm{s}$ .

Step 1. During the push the skaters exert equal and opposite forces on each other at every instant — equation (1) — for the same contact time, so they receive equal and opposite impulses. Compute the heavier skater's momentum: $p_B = m_B v_B = (72)(0.90) = 64.8\ \mathrm{kg\,m/s}$ .

Step 2. The lighter skater carries the same magnitude of momentum the other way: $p_A = 64.8\ \mathrm{kg\,m/s}$ , so $v_A = p_A/m_A = 64.8/48 = \boxed{1.35\ \mathrm{m/s}}$ — faster in exact inverse proportion to the mass ratio, $v_A/v_B = m_B/m_A = 72/48 = 1.5$ .

Step 3. Check the ledger: $\mathbf{p}_{\text{tot}} = (48)(1.35) - (72)(0.90) = 64.8 - 64.8 = \mathbf{0}$ , exactly as it began. The system started at rest and no external horizontal force acted, so total momentum had to stay zero — equation (3) in action.

Step 4. The average force on each skater is the impulse over the contact time: $\bar F = 64.8/0.60 = \boxed{108\ \mathrm{N}}$ . Both skaters feel the same $108\ \mathrm{N}$ — the big skater enjoys no discount; only the response to the force (the acceleration) differs.

2 · The tension in a tow bar (exam level)

A $1200\ \mathrm{kg}$ car tows a $400\ \mathrm{kg}$ trailer along a straight road, accelerating at $1.5\ \mathrm{m/s^2}$ . Rolling and air resistance total $300\ \mathrm{N}$ on the car and $200\ \mathrm{N}$ on the trailer. Find the driving force and the force in the tow bar, then verify the bar force from both ends.

Step 1. Treat car + trailer as one system — legal because the bar forces are an internal Third-Law pair and cancel by (4). The centre-of-mass theorem (5) gives $F_{\text{drive}} = (m_c+m_t)a + R_c + R_t = (1600)(1.5) + 500 = \boxed{2900\ \mathrm{N}}$ .

Step 2. Now isolate the trailer. It is pulled forward only by the bar: $T - R_t = m_t a$ , so $T = (400)(1.5) + 200 = \boxed{800\ \mathrm{N}}$ .

Step 3. Consistency check from the car's own diagram, where the Third Law says the bar (via the trailer) pulls backward with the same $800\ \mathrm{N}$ : $F_{\text{drive}} - R_c - T = 2900 - 300 - 800 = 1800\ \mathrm{N} = (1200)(1.5)$ ✓. The same magnitude $T$ appears in both diagrams with opposite signs — that is the pair.

Step 4. Read the result physically: the hitch carries $800\ \mathrm{N}$ whichever end you stand at. A strain gauge in the bar would read one number, not two — a Third-Law pair is a single interaction with two addresses.

Figure

The tow bar carries one force with two addresses

Origins · 1 minIsaac Newton and how this was discovered

Isaac Newton

1642–1727 · Cambridge

Newton composed the Principia in an eighteen-month surge of work, prodded and financed by Edmond Halley, and stated the Third Law as the last of his three axioms — Lex III — illustrating it with a finger pressing a stone ("the finger is also pressed by the stone") and a horse drawing a rope. Of the three laws it was the most original: the first two refined ideas circulating since Galileo and Descartes, but the systematic pairing of forces, and the momentum bookkeeping it enables, was Newton's own. He needed it, too — without the Third Law the Earth–Moon system in Book III would not conserve momentum, and universal gravitation would not hang together. Master of the Mint in his later decades, he died in London in 1727 and was buried in Westminster Abbey.

History

1687

— Newton publishes the Principia; Lex III pairs every action with an equal, opposite reaction, completing the axioms of mechanics.

1743

— d'Alembert's Traité de dynamique recasts dynamics so that constraint forces cancel in pairs — the Third Law absorbed into method.

1750

— Euler extends Newton's laws to rigid bodies and continua, with pairwise cancellation of internal forces as the working backbone.

1884

— Poynting quantifies the energy and momentum flowing in the electromagnetic field — the ledger entry that rescues conservation where moving charges violate the Third Law.

1918

— Noether proves momentum conservation follows from translation symmetry of space: the modern foundation beneath the Third Law.

Going further

References. Morin, Introduction to Classical Mechanics, Ch. 3 · Goldstein, Classical Mechanics (3rd ed.), Ch. 1, §1.1–1.2 · Landau & Lifshitz, Mechanics (Vol. 1), §3 · Taylor, Classical Mechanics, Ch. 1, §1.5 · Newton, Philosophiæ Naturalis Principia Mathematica (1687), Axioms, Lex III Key results in this article: the momentum-conservation derivation gives $\mathrm{d}\mathbf{p}_{\text{tot}}/\mathrm{d}t=\mathbf{F}_{AB}+\mathbf{F}_{BA}=0$ ; the stacked-block results are $N_{12}=m_1 g$ and $N_{2g}=(m_1+m_2)g$ from static equilibrium; the skater speed ratio is $v_A/v_B=m_B/m_A$ from equal impulses; strong-form torque cancellation requires collinearity. The magnetic-force counterexample is resolved by the field carrying momentum $\int\varepsilon_0\,\mathbf{E}\times\mathbf{B}\,\mathrm{d}V$ . Worked numbers: skaters $v_A = 72\times0.90/48 = 1.35$ m/s, impulse $64.8$ kg·m/s, mean force $108$ N; tow bar $T = 400\times1.5+200 = 800$ N, drive force $2900$ N, car check $2900-300-800 = 1800 = 1200\times1.5$ .

Prerequisites. Newton's Second Law · Newton's First Law and Inertial Frames · Vectors in Kinematics and Coordinate Systems. Leads to. Normal Force · Friction Forces · Gravitational Force Near Earth's Surface.

If every action has an equal and opposite reaction, how does anything ever accelerate?

Because the two forces of a pair act on different bodies, they can never cancel. To apply $\mathbf{F}=m\mathbf{a}$ to a body you sum only the forces on that body — the reaction lives in the other body's free-body diagram. When a horse pulls a cart, the cart's backward pull acts on the horse, not on the cart; whether the cart accelerates depends solely on the forces on the cart. The horse moves because the ground's forward push on its hooves (the reaction to its own backward shove on the ground) exceeds the cart's backward pull on it. Nothing requires the forces on a single body to balance — that only happens in equilibrium, which is a Second-Law statement, not a Third-Law one.

Isn't the normal force on a resting book just the reaction to its weight?

No — apply the pairing test: a reaction is the same interaction with the bodies swapped. The book's weight is "Earth pulls book, gravitationally"; its reaction is "book pulls Earth, gravitationally" — the book genuinely tugs the planet upward. The normal force is a different interaction entirely, "table pushes book, by contact," whose reaction is "book pushes table." The two forces on the book are equal only because the book happens to have $a=0$ . Put the whole scene in a dropping elevator and $N$ falls below $mg$ instantly, while both true pairs stay perfectly matched throughout. Same-body forces that balance are a Second-Law coincidence of equilibrium, never a Third-Law pair.

When a truck hits a small car, doesn't the truck exert the bigger force?

No. Equation (1) has no loophole for size: at every instant of contact the truck pushes on the car exactly as hard as the car pushes back. What differs is the response, $a=F/m$ — same force, wildly different accelerations. The skaters make it quantitative: both feel the same $108\ \mathrm{N}$ , yet the 48 kg skater accelerates 1.5 times as much as the 72 kg one. In a crash, the car's smaller mass turns the shared force into a violent acceleration, and its structure tolerates less of it, so the damage is lopsided even though the forces are not. Force is always symmetric; consequences never have to be.

Does the Earth really accelerate upward toward a falling apple?

Yes — and equation (3) tells you by how much. As the apple gains downward momentum, the Earth gains exactly equal upward momentum, and their common centre of mass never moves. The accelerations scale inversely with mass: while the apple picks up $9.8\ \mathrm{m/s^2}$ , the Earth responds at roughly $10^{-25}\ \mathrm{m/s^2}$ — real, but buried far below anything measurable. The effect stops being academic when the masses are comparable: the Moon and Earth visibly orbit their common centre of mass, and the reflex wobble a planet's pull imposes on its star — the same gravity pair at stellar scale — is how hundreds of exoplanets have been discovered.

How can a rocket accelerate in empty space with nothing to push against?

It pushes against its own exhaust. The engine hurls gas backward, and by equation (1) the gas, in the very act of being accelerated, pushes the rocket forward with equal force — the pair is complete in itself, no road, air, or launch pad required. The momentum ledger (3) says the same thing: every kilogram of exhaust thrown back at high speed carries backward momentum, and the rocket pockets the equal forward amount. Air would only add drag; rockets actually work better in vacuum. The two recoiling skaters are the same physics — a rocket is a machine that pushes off from a continuous stream of tiny "skaters," its own gas molecules.

Why do we need a separate strong form — isn't equal and opposite enough?

Enough for linear momentum, not for angular. In the torque derivation, the weak form collapses a pair's torque to $(\mathbf{r}_A-\mathbf{r}_B)\times\mathbf{F}_{BA}$ — and that is not automatically zero. Two equal and opposite forces offset sideways form a couple, and a couple has net torque; if internal forces could form couples, an isolated system could spin itself up from nothing. Collinearity — the force lying along the line joining the bodies, as in equation (2) — makes the cross product a product of parallel vectors, hence zero. Gravity and electrostatics obey the strong form, which is why the weak form buys you conserved linear momentum while the strong form is needed to conserve angular momentum as well.

If magnetic forces break the Third Law, why isn't momentum conservation dead?

Because the accounting was incomplete, not wrong. Two charges moving perpendicular to their separation exert magnetic forces that fail even the weak form, so the particles' total momentum genuinely changes. The missing momentum sits in the electromagnetic field itself, which carries momentum $\int\varepsilon_0\,\mathbf{E}\times\mathbf{B}\,\mathrm{d}V$ ; add the field to the ledger and conservation is exact again. This is no bookkeeping trick — field momentum arrives as real radiation pressure, the push optical tweezers use to hold living cells. The deeper lesson: the Third Law is the instantaneous, pairwise expression of momentum conservation, and the conservation law — guaranteed by translation symmetry via Noether's theorem — is the part that survives when "force pairs" no longer do.

Both skaters feel the same force, so why don't they end up with the same kinetic energy?

The Third Law equalizes impulses, not energies. Each skater receives $64.8\ \mathrm{kg\,m/s}$ , but kinetic energy is $p^2/2m$ : the 48 kg skater ends with about $43.7\ \mathrm{J}$ , the 72 kg skater with only $29.2\ \mathrm{J}$ . The asymmetry is geometric — both feel the same force for the same time, but the lighter skater accelerates away faster, so her centre of mass covers more distance while the force acts, and energy transfer scales with force times distance. The energy itself is chemical, supplied by their muscles during the push, and it splits in inverse proportion to mass. Momentum symmetry is the Third Law's guarantee; how the energy divides is a separate question the law says nothing about.

Are action and reaction really simultaneous — doesn't that clash with relativity?

It does, strictly speaking. For contact forces the two ends of the interaction sit at the same place, so simultaneity is harmless. But for two charges a distance $r$ apart, the force on A now reflects what B was doing a time $r/c$ ago — the interaction is retarded, and "equal and opposite at every instant" cannot even be stated cleanly across the gap. The instantaneous Third Law is the $v\ll c$ , negligible-retardation approximation, as the limiting-cases table notes. What survives exactly is the conservation law: momentum leaving one particle travels through the field, which carries momentum of its own, and is delivered to the other particle later. Relativity demotes the Third Law and promotes what it was encoding all along — momentum conservation from translation symmetry.

Newton's Third Law

Where this lives today

The physical picture

Pairs, not balances

From pairs to conserved momentum

Angular momentum needs the strong form

The impulse view — collisions in one line

Chains of pairs — the stacked blocks

Limiting cases

Worked examples

1 · Two skaters push apart (confidence builder)

2 · The tension in a tow bar (exam level)

Going further

Before this

Where this leads