Working note · analytical mechanics · part X (capstone of the series)
Principle of least action: one idea from which all physics follows
Newton's three laws, conservation of energy and momentum, refraction of light, Maxwell's and Schrödinger's equations — all of these can be derived from one line: δS = 0.
In parts I and II of this series, we already twice encountered a hint of a deeper level: both with Newton and with conservation laws, there was a §6-teaser about the principle of least action and Noether's theorem. Here is a full analysis: what is action S, why nature 'chooses' the path with its minimum, how Newton's equations are derived from it, and how Noether's theorem of 1918 explains WHY conservation laws exist at all — not as isolated happy coincidences, but as a direct consequence of the symmetries of space and time.
One idea instead of three laws
Until now in this series, physics was built 'bottom-up': forces, accelerations, trajectories — calculate step by step. The principle of least action offers a 'top-down' view: not calculating the trajectory step by step, but immediately asking — which of ALL CONCEIVABLE trajectories between point A and point B is the real one? Answer: the one for which a special quantity, action S, is minimal (more precisely — stationary, but for most everyday problems it is a minimum).
| Approach | What we calculate |
|---|---|
| Newton (parts I-IX) | force at each moment → acceleration → next position, step by step |
| Least action | a single number S for EACH entire trajectory as a whole → choose the minimum |
Both approaches give the SAME answer for classical mechanics (§3) — they are not competing theories, but two languages describing the same physics. But the language of action turns out to be more universal: it works equally well where the concept of 'force' is inconvenient or nonexistent — in optics, electromagnetism, quantum mechanics, general relativity.
What is action
Formulation. Action is a number calculated for the entire trajectory of motion:
Let's check the simplest case — a free particle (no forces, U = 0) that must travel distance x = 8 m in time T = 4 s. Compare the action for two different trajectory 'candidates':
| Candidate | Description | Action S |
|---|---|---|
| A (green) | uniform motion, v = 2 m/s all the time | 16 J·s |
| B (orange, dashed) | stays still for 2s, then suddenly accelerates to 4 m/s | 32 J·s |
Calculation for A (m = 2 kg): L = K = ½·2·2² = 4 J constant, S₁ = 4·4 = 16 J·s. For B: first 2 s K = 0, next 2 s K = ½·2·4² = 16 J, S₂ = 0·2+16·2 = 32 J·s. Uniform motion wins — exactly what Newton's first law (part I, §2) predicts: a free body moves with constant velocity. The principle of least action not only AGREES with this — it generates it.
From the principle — back to Newton
The condition of minimal action (calculus of variations, δS = 0) gives a universal recipe — Euler-Lagrange equation:
Substitute L = ½mṲ² − U(q) for an ordinary particle (q — coordinate, Ṳ — velocity). ∂L/∂Ṳ = mṲ = p (momentum), ∂L/∂q = −∂U/∂q = F (force — minus gradient of potential energy). Equation (1) becomes:
This is not a coincidence or a trick — the Euler-Lagrange equation is identical to F = ma for ANY potential U. All of classical mechanics is a special case of a single variational principle.
Noether's theorem: symmetries give birth to laws
In 1918, Emmy Noether proved one of the deepest theorems of physics: to each continuous symmetry of the action S corresponds its own conserved quantity. Not individually discovered empirical facts, but a single source for all conservation laws:
| Symmetry of the action | Conserved quantity |
|---|---|
| homogeneity of time (shift t → t+Δt changes nothing) | energy (part II, §3) |
| homogeneity of space (shift x → x+Δx changes nothing) | momentum (part II, §2) |
| isotropy of space (rotation by any angle changes nothing) | angular momentum (part II, §4; part VII, §4) |
This flips the usual order of cause and effect: not 'it just so happens that energy is conserved', but 'time is homogeneous, therefore energy MUST be conserved' — a rigorous mathematical consequence, not an empirical observation. More abstract symmetries (not of spacetime, but of internal 'phase' degrees of freedom of fields) similarly give rise to conservation of electric charge and other quantum numbers — the theorem works far beyond mechanics.
Emmy Noether derived the theorem in Göttingen, working on the problem of energy conservation in general relativity at the request of Hilbert and Klein. Einstein called her result 'a monument of mathematical thought'. At the same time, due to discrimination against women in science at that time, Noether for many years could not obtain a paid professorship — she gave lectures formally 'under the name' of Hilbert.
Running example: why light refracts
Fermat's principle (a special case of the principle of least action for optics, from the 17th century): light propagates along the path requiring the LEAST time — not necessarily the shortest distance.
Imagine a lifeguard on the beach who has to run to a drowning swimmer — he can run fast on sand, swim slower in water. The shortest PATH (straight line) is not the fastest: it is more advantageous to run a little more on the sand to shorten the slow segment in the water. The optimal entry point into the water is not where the straight line is, but a little 'at an angle'. Light behaves exactly the same at the air/water boundary, and the precise calculation of the optimal angle yields exactly Snell's law of refraction.
The 'lifeguard on the beach' problem is mathematically IDENTICAL to the principle of least action for a particle moving from a region with one potential energy to a region with another (refraction of a charged particle's trajectory at a field boundary) — optics and particle mechanics turn out to be two disguised versions of the same variational problem. A detailed analysis of the law of refraction with formulas and numbers is a separate note in the series, the next one on optics.
On the site, these are separate laws if you want to go deeper:
Where this leads: sum over all paths
Pierre Louis Maupertuis first formulated the principle of least action in the 1740s as teleological — as if nature 'knows' the best path in advance and strives towards a goal, almost a design. The modern reading, after the work of Richard Feynman (1940s), is quite different and no less amazing: the particle does not choose a path in advance, but in a quantum sense traverses ALL conceivable paths simultaneously, each with its own complex 'phase' amplitude. The paths add up (interfere) in such a way that for most of them the contributions cancel each other out, and in the macroscopic limit practically only the path with extremal (minimal) action survives — hence Newton's classical mechanics as a limiting case.
Schrödinger's equation (part VI) and Maxwell's equations (part IV) — both are derived from a suitably chosen action S through exactly the same mathematics δS = 0 as the ordinary F = ma here. That is the meaning of the phrase 'one idea from which all physics follows': not a poetic exaggeration, but a working fact of modern theoretical physics — from school mechanics to quantum field theory and string theory.
Home experiment
Bend a wire (e.g., from a hanger) into any non-planar shape — for example, two rings connected crosswise with a bridge — and dip it into a soap solution (water + liquid dish soap, a little glycerin for strength).
What to notice the film itself 'finds' the surface of MINIMAL area stretched on the frame (minimal surface) — surface tension at each point acts to reduce the total energy (proportional to area), and the result is the exact solution of a variational problem 'minimize a functional' — the same mathematical idea as minimizing action, just for a different physical quantity.
Make two tracks of equal length between the same start and end points (height difference is the same) — one straight (inclined plane), the other curved with a concave arc near the start. Release two identical balls at the same time.
What to notice the ball on the concave path usually reaches the finish faster than the straight one, although its path is longer — historically, it was this problem (brachistochrone, 'curve of shortest time') in the 17th century that prompted mathematicians to create the calculus of variations, the very apparatus on which the principle of least action stands.
Problems for check
First think for yourself, then open the solution. The solution everywhere follows the scheme: Given → Law → Solution → Answer.
1. A free particle of mass 4 kg must travel 12 m in 3 s. Find the action S for uniform motion.
Given m = 4 kg, distance = 12 m, T = 3 s, U = 0 (free particle).
Law §2: S = ∫L dt, L = K for a free particle, with uniform motion K is constant.
Solution v = 12/3 = 4 m/s; K = ½·4·4² = 32 J; S = 32·3 = 96 J·s.
Answer 96 J·s — and this is the MINIMAL possible action among all trajectories with the same start/end (§2).
2. Which symmetry of the action, according to Noether's theorem, is responsible for the conservation of momentum?
Law Noether's theorem (§4).
Solution momentum is conserved if the action does not change under a spatial shift of the system — that is, the laws of physics are the same at any point in space (there is no 'special place' in the Universe).
Answer homogeneity of space.
3. Why does Fermat's principle (light follows the path of least TIME) not mean that light always follows the shortest DISTANCE?
Law §5 — 'lifeguard on the beach' problem.
Solution time equals distance divided by speed, and the speed of light is different in different media (slower in water/glass than in air) — minimization of TIME sometimes requires traveling a GREATER distance in a fast medium to shorten the path in the slow one, just as it is more advantageous for the lifeguard to run an extra stretch on the sand.
Answer because the speed of light depends on the medium — shortest time and shortest distance coincide only in a homogeneous medium.
4. The Euler-Lagrange equation, applied to L = ½m·q̇² − U(q), gives F = ma. What does the symbol q̇ physically represent?
Law notations of calculus of variations (§3).
Solution the dot over q is the standard notation for the time derivative (adopted by Newton); q̇ = dq/dt.
Answer velocity (time derivative of coordinate q).
5. Why is Noether's theorem not directly applied to discrete symmetries (for example, mirror parity — spatial reflection)?
Law limitations of Noether's theorem (§4, key_problems of the law).
Solution Noether's theorem requires a CONTINUOUS symmetry — a transformation that can be performed by an arbitrarily small amount (shift by 0.0001 m, rotation by 0.0001°). Reflection (parity) is a discrete operation: it cannot be done 'a little' — either it is reflected or not, so the classical formulation of the theorem is not directly applicable to it (although in quantum field theory there are separate, more subtle results about discrete symmetries, for example the CPT theorem).
Answer parity is a discrete, not continuous symmetry — it lies outside the scope of the original Noether's theorem.
Typo, wrong formula, number doesn't add up, clumsily explained — write to us: feedback@bridge42worlds.org. Every message is actually read, analyzed, and corrections are made based on it — it's not 'into the void'.