Action-reaction pairs

In brief

Stand on a skateboard next to a wall and shove the wall hard, and you roll backward even though nothing pulled or dragged you, the wall never touched you except during that one push. Something pushed you back, and the only candidate is the wall itself, pushing on you exactly as you pushed on it. This is not a special property of walls; it is a fact about every interaction between two objects. Whenever object A exerts a force on object B, object B exerts a force of equal size and exactly opposite direction back on object A, at the same instant. This single fact resolves a puzzle that looks, at first, like magic: how a rocket in the emptiness of space, with nothing around it to push against, can still accelerate forward by throwing exhaust out the back.

The full treatment

First look: two hands, one squeeze

Press your right palm against your left palm as hard as you can. You feel pressure on both hands, equally. Push harder with the right hand and, without trying, the pressure on the left hand rises to match, you cannot make one hand push on the other without the other pushing back with the same strength. Try to violate this: attempt to push your right palm against your left with more force on one side than the other. It cannot be done, because "pushing on the left hand" and "pushing on the right hand" here are not two independent events, they are two descriptions of the same single squeezing interaction between your two hands.

Building the idea: forces always come from a shared interaction

Recall that a force is not something an object possesses, it is what happens between two objects during an interaction. This reframing, developed in the entry on force as interaction, already contains the seed of the pairing. If "force of A on B" is just one name for an interaction between A and B, then "force of B on A" is the same interaction viewed from the other side, not a separate coincidence layered on top. What Newton's third law adds, and what is not automatic from wording alone, is the precise, quantitative claim: these two forces are equal in magnitude and opposite in direction, at every instant, for every kind of interaction, contact, friction, gravity, magnetism, without exception. This was an empirical discovery, confirmed by careful collision experiments, not a fact you can derive from grammar alone.

The formal model: naming the pair correctly

Write the force exerted by object A on object B as F(A on B), and the force exerted by object B on object A as F(B on A). Newton's third law states F(A on B) equals negative F(B on A): equal magnitude, opposite direction. A critical detail, and the single most common source of confusion, is that these two forces act on two different objects, one on A and one on B, and for that reason they never cancel each other out or produce equilibrium in either object considered alone. This distinguishes an action-reaction pair from an equilibrium pair, such as gravity pulling a book down and the table pushing it up, which are two different forces acting on the same single object (the book) and happen to sum to zero. An action-reaction pair is always two forces on two different objects; an equilibrium pair is often two forces on the same object. Confusing the two is the single most common error made when first learning this law.

Mechanism: why you can push off nothing solid

Consider the rocket puzzle directly. A rocket in deep space, far from any planet, ejects burned propellant gas backward at high speed. The rocket engine exerts a force on the gas, accelerating it backward and out of the nozzle. By the third law, the gas exerts an equal and opposite force on the rocket, forward, at that same instant, regardless of whether there is a planet, a wall, or any other outside object nearby. The rocket does not need "something to push against" in the sense of a fixed external object; it only needs a second object to interact with, and the exhaust gas itself is that second object. This is why rockets work in the vacuum of space, where propellers and wheels, which do rely on pushing against an external medium or surface, cannot function at all. Swimming and walking work by the same logic on a smaller scale: a swimmer pushes water backward and the water pushes the swimmer forward; a walker pushes the ground backward (through friction) and the ground pushes the walker forward, in both cases using a genuine second object as the partner in the interaction, not "nothing."

A worked check: recoil

A gun of mass M, initially at rest, fires a bullet of mass m forward. During the brief firing interaction, the gun exerts a forward force on the bullet, and by the third law, the bullet exerts an equal and opposite (backward) force on the gun, for the same duration. Equal force over equal time on each object means each receives an equal and opposite impulse, and since impulse changes momentum, the gun recoils backward with exactly enough momentum to match the bullet's forward momentum, which is why a fired rifle kicks the shooter's shoulder. Nothing about the surrounding air or ground is needed to explain the kick; it follows entirely from the bullet-gun pair.

Lineage

Newton stated the third law in the Principia Mathematica of 1687 as one of three foundational laws of motion, alongside the law of inertia and the force-mass-acceleration relation, phrasing it as "to every action there is always opposed an equal reaction: or, the mutual actions of two bodies upon each other are always equal, and directed to contrary parts." The law resolved older confusions about collisions and recoil that had puzzled natural philosophers such as Descartes, and it was quickly recognized as necessary for consistency with the conservation of momentum, itself independently studied by Christiaan Huygens in collision experiments in the same period. The rocket application, while implicit in the law from the start, became a celebrated illustration once spaceflight made "how can you push off nothing" a genuinely pressing engineering question in the twentieth century.

The strongest case for it

The law is confirmed continuously by every measured collision, every rocket launch, and every recoil event ever tested, and it underlies the conservation of momentum, one of the most reliable principles in all of physics, which follows directly from action-reaction pairs summed over an entire isolated system. Its predictive reach is enormous: it explains why swimming, walking, and rowing work, why a fired projectile recoils its launcher, why a rocket needs no external medium, and why two skaters pushing off each other fly apart with momenta that are equal and opposite. Engineers rely on it explicitly, every propulsion system, from oars to jet engines to rocket motors, is designed around identifying the second object being pushed and calculating the reaction it produces.

The strongest case against it

The clean pairing describes the fundamental interaction correctly, but it is easy to misapply. The most common error is searching for the reaction force on the wrong object, for instance, expecting the "reaction" to a car's engine force to appear on the car itself (it does not; the paired force is on the road, via friction, and it is a separate pair, road pushing car forward and car pushing road backward, that actually explains the car's own acceleration). A second common error is assuming action-reaction pairs cancel and therefore nothing can ever accelerate, but since the two forces act on two different objects, each object separately experiences its own single force and accelerates according to its own mass, the pairing never implies zero net force on either object. A genuine boundary of the classical law, rarely relevant at everyday scale, is that at the quantum level, "instantaneous" equal and opposite forces at a distance conflict with the finite speed of any real interaction (no influence travels faster than light), which is why field-based reformulations, where the field itself carries momentum during the brief transit time, replace the strict instantaneous version in relativistic treatments.

Where it stands now

Newton's third law, understood correctly as a statement about forces on two distinct objects arising from one interaction, remains exact and universally applied throughout classical mechanics and engineering. Its momentum-conservation consequence is one of the most stringently tested principles in physics, verified from collisions in particle accelerators to the orbits of binary stars. The only refinement modern physics adds is at relativistic and field-theoretic scales, where momentum carried by fields during a finite transit time completes the accounting; this refines, rather than overturns, the everyday pairing this entry describes.

Test yourself

An astronaut is stranded, motionless relative to her spacecraft, floating in open space with no rope, no thruster pack, and nothing solid nearby to grab. She has only her space suit, a wrench, and her own two hands. Using only the idea of action-reaction pairs, describe a method by which she can set herself moving toward the spacecraft, and identify precisely which two objects form the interacting pair in your method. Then explain why the same method would fail if she instead tried to move by only waving her arms through the vacuum without releasing or throwing anything away from her body.