At the heart of modern physics is a quiet pact between symmetry and what can’t change. By starting from the principle of least action, introducing the Lagrangian, and then applying calculus of variations, Noether’s theorem shows that every continuous symmetry
Last week, I introduced Emmy Noether—an extraordinary figure in mathematics and physics. The core promise of her theorem is deceptively simple: for every continuous symmetry of a system. there is a conserved quantity. It’s one of those ideas that makes the universe feel less like chaos and more like a logic you can lean on.
But how does that reasoning actually work?
To get there, the story has to start with how physicists describe motion. In many high school physics problems—like determining the orbit of a planet around a star or the trajectory of a ball—students begin with force equations. The gravitational force between two bodies. for instance. is set equal to mass times acceleration of the planet. and from there an equation of motion tells you when and where the object goes.
In college, physics students learn a different, more flexible approach: energy. The force method and the energy method are equivalent. producing the same results. but the energy approach is often more practical—and crucially. easier to generalize. That generalization is what makes Noether’s theorem possible.
The energy method rests on an idea that sounds almost human: the principle of least action. Nature. in this view. is “lazy.” When a system transitions from one state to another—for example. when a ball flies through the air and then lands on the ground—it follows the path of least effort. In physics, that effort is called action. The insight connects back to Fermat’s principle. where light rays choose the shortest path to a destination. and then expands the pattern outward as if other systems are doing the same.
By taking this principle and applying calculation, you can derive equations of motion—such as the orbits of planets around the sun.
To carry out this program, physicists introduce the Lagrangian, a fundamental function for dynamic systems. If you want to fully characterize a thrown ball, you need its velocity and position at every instant. Keeping track of everything at once can be overwhelming: those quantities are described by a six-dimensional vector—three coordinates for position and three for velocity—changing as time passes. The Lagrangian compresses that complexity into a single scalar quantity that tracks the system over time.
When the Lagrangian’s value changes, it reflects movement within the system. The action is tied to it directly: action is given by the sum (more precisely. the integral) of the Lagrangian at each individual instant. In other words, action assigns a numerical value to every possible trajectory. The correct motion of the physical system then corresponds to the principle of least action—the trajectory that makes the action smallest. or at least stationary. compared with nearby alternatives.
In calculus terms, students are used to finding highest and lowest points of a function within an interval. Those peaks and valleys are extrema, found by differentiating and setting the result equal to zero. But action isn’t just a normal function. It’s a different creature: a functional. The action integrates the Lagrangian over time, and the Lagrangian itself is built from time-dependent functions like velocity and position. So the mathematics has to be handled differently.
The route forward is the calculus of variations. The method feels familiar—tweak the possible trajectories and look for where the action changes the least—but the objects being varied aren’t inputs to a function; they’re entire paths. By doing this carefully. you obtain equations that correspond to the equations of motion of the system. including examples like the orbits of planets.
Then comes Noether’s move—the one that turns all of this machinery into a bridge between symmetry and conservation.
The Lagrangian, it turns out, lets you read symmetries directly. If you apply a symmetry transformation—like shifting the x coordinates—and the Lagrangian L stays unchanged. you’ve found a symmetry. A simple example makes the idea concrete. Imagine two spheres moving toward each other along the x axis and colliding. In that setup, the Lagrangian depends only on their distance. Written as s1 − s2 = q. q is the generalized coordinate. s1 is the position of sphere one. and s2 is the position of sphere two.
Now shift both spheres by the same distance α. The relative separation stays the same because (s1 + α) − (s2 + α) = q. The Lagrangian remains unchanged, and the system is symmetric with respect to translation.
Noether investigated what happens when a Lagrangian changes as a variable—like time or position—is varied by a parameter α. The change in L is analyzed by taking the derivative of the Lagrangian with respect to α. If that derivative is zero—meaning L doesn’t change under the symmetry transformation—then something striking follows.
Using properties of the Lagrangian and a set of transformations. the derivative of L with respect to α. expressed as (∂L/∂α). becomes the derivative of a new expression Q with respect to time. written as dQ/dt. And that derivative is also zero. Q does not change over time, so it becomes a conserved quantity.
That’s the heart of Noether’s theorem: every continuous symmetry corresponds to a conserved quantity, and Noether’s framework even gives a formula for calculating it.
The chain of ideas—from forces to energy. from energy to least action. from least action to the Lagrangian. and from the behavior of that Lagrangian under shifts—to conservation laws. isn’t just elegant. It makes a promise about the universe’s accounting: when a system has a continuous symmetry. there’s a number somewhere inside the equations that refuses to drift.
This article originally appeared in Spektrum der Wissenschaft and was reproduced with permission. It was translated from the original German version with the assistance of artificial intelligence and reviewed by our editors.
Noether's theorem Emmy Noether symmetry conserved quantities Lagrangian principle of least action calculus of variations physics foundations