Skip to content

Continuous Symmetries and Conservation Laws

Continuous transformations connect a finite operation to an infinitesimal generator. When that transformation also preserves the dynamics, its generator is conserved. This chapter develops the distinction: momentum generates translations even in a trap, but the trap prevents momentum conservation. It also explains why a conserved observable can have a spread of outcomes and why explicitly time-dependent observables need a more general test.

Helpful background. Unitary symmetries give the finite transformation law; commutators give its infinitesimal action.

Question Read next
Which continuous families have a fixed generator? One-parameter unitary groups
How do we find and use the generator? Generators and infinitesimal transformations
When does a symmetry give conservation? Commutators and conservation laws and the quantum Noether principle
What if an observable depends explicitly on time? Constants of motion
Which labels and matrix elements are restricted? Good quantum numbers and selection rules
What survives a perturbation? Broken symmetry

A group law is stronger than a smooth family

Section titled “A group law is stronger than a smooth family”

A strongly continuous one-parameter unitary group obeys U(0)=IU(0)=I and U(α+β)=U(α)U(β)U(\alpha+\beta)=U(\alpha)U(\beta). Stone’s theorem gives a self-adjoint GG such that

U(α)=e−iαG/ℏ.U(\alpha)=e^{-i\alpha G/\hbar}.

Continuity means U(α)∣ψ⟩U(\alpha)|\psi\rangle varies continuously in norm for every vector. Differentiability is stronger: the derivative at zero exists for vectors in the generator’s domain. On that domain,

U(δα)∣ψ⟩=∣ψ⟩−iδαℏG∣ψ⟩+o(δα).U(\delta\alpha)|\psi\rangle =|\psi\rangle-\frac{i\delta\alpha}{\hbar}G|\psi\rangle +o(\delta\alpha).

A general smooth family of unitaries need not have one fixed generator or obey the additive group law. Time evolution under a varying Hamiltonian is an important example: its two-time propagators compose, but generally do not depend only on elapsed time.

The product αG\alpha G has units of action. For translations, α\alpha is length and GG is momentum; for rotations, the angle is dimensionless and GG has angular-momentum units. See Stone’s theorem for the precise domain statement.

For finite matrices, or on vectors where the required unbounded products and derivatives exist,

UAU†=A−iδαℏ[G,A]+O(δα2),U†AU=A+iδαℏ[G,A]+O(δα2).\begin{aligned} UAU^\dagger&=A-\frac{i\delta\alpha}{\hbar}[G,A]+O(\delta\alpha^2),\\ U^\dagger A U&=A+\frac{i\delta\alpha}{\hbar}[G,A]+O(\delta\alpha^2). \end{aligned}

The opposite signs answer opposite transformation questions. The second expression computes a fixed observable’s expectation in the transformed state. For instance, with T(a)=e−iaP/ℏT(a)=e^{-iaP/\hbar}, T†XT=X+aIT^\dagger XT=X+aI, so the wavepacket moves right by aa. The first-order formula is not a substitute for the finite exponential when the parameter change is large.

Conservation includes explicit time dependence

Section titled “Conservation includes explicit time dependence”

For a possibly time-dependent Schrödinger-picture observable A(t)A(t),

ddt⟨A⟩=iℏ⟨[H,A]⟩+⟨∂A∂t⟩.\frac{d}{dt}\langle A\rangle = \frac{i}{\hbar}\langle[H,A]\rangle + \left\langle \frac{\partial A}{\partial t} \right\rangle.

The corresponding operator criterion for a constant of motion is

∂A∂t+iℏ[H,A]=0.\frac{\partial A}{\partial t} + \frac{i}{\hbar}[H,A] =0.

Thus [H,A]=0[H,A]=0 is sufficient only when AA has no explicit time dependence. Conversely, an explicitly time-dependent A(t)A(t) can still be conserved when its explicit derivative cancels its commutator term.

For a continuous unitary symmetry generated by GG, the ordinary quantum Noether pattern is

U(α)HU†(α)=H⇓[G,H]=0⇓dGHdt=0,\begin{aligned} U(\alpha)HU^\dagger(\alpha)&=H \\ &\Downarrow \\ [G,H]&=0 \\ &\Downarrow \\ \frac{dG_H}{dt}&=0, \end{aligned}

where the last implication assumes ∂G/∂t=0\partial G/\partial t=0.

For unbounded operators, these differentiations require suitable states and domains; strong commutation of spectral measures supplies the finite symmetry statement. The commutator argument alone is not the full field-theory theorem. Classical Noether theory begins with an action and treats boundary terms. Field theory produces a local current jμj^\mu and a continuity equation ∂μjμ=0\partial_\mu j^\mu=0 before a conserved charge is obtained. The ordinary quantum-mechanical chapter stays with Hilbert-space transformations, Hamiltonians, generators, and commutators. The field-theory bridge begins at From Quantum Generators to Noether Currents.

If a time-independent observable strongly commutes with the Hamiltonian, all its spectral probabilities are preserved by time evolution. A free wavepacket may therefore have a broad, unchanged momentum distribution. By contrast, an expectation value can happen to be stationary in one special state even when the operator is not conserved: the harmonic oscillator has ⟨P⟩=0\langle P\rangle=0 in an energy eigenstate, although [H,P]≠0[H,P]\ne0.

Good quantum numbers label compatible spectral sectors. In a discrete-spectrum problem, [H,Q]=0[H,Q]=0 with the appropriate operator assumptions lets energy states be chosen with a QQ label. A second proposed label must also be compatible with the first. Rotationally invariant dynamics conserves each JiJ_i, but the noncommuting components cannot all label a basis sharply; the usual choice is J2,JzJ^2,J_z.

Let QQ generate a continuous symmetry, and suppose

Q∣i⟩=qi∣i⟩,Q∣f⟩=qf∣f⟩.Q|i\rangle=q_i|i\rangle, \qquad Q|f\rangle=q_f|f\rangle.

If a transition operator TT has a definite commutator with QQ,

[Q,T]=qTT,[Q,T]=q_T T,

then

(qf−qi)⟨f∣T∣i⟩=⟨f∣[Q,T]∣i⟩=qT⟨f∣T∣i⟩.\begin{aligned} (q_f-q_i)\langle f|T|i\rangle &= \langle f|[Q,T]|i\rangle \\ &= q_T\langle f|T|i\rangle. \end{aligned}

A nonzero matrix element therefore requires

qf−qi=qT.q_f-q_i=q_T.

This is a symmetry constraint, not a dynamical prediction of the transition rate. Satisfying the condition is generally necessary, not sufficient: radial integrals, additional symmetries, kinematics, and accidental cancellations can still make the matrix element vanish.

If a perturbation mixes the old symmetry sectors, the old rule may become approximate. Small coupling is not enough by itself to make this mixing small: it must be compared with the relevant energy gaps, and near-degenerate sectors can mix strongly. See good quantum numbers for a two-level example.

  1. A particle has H=P2/(2m)+12mω2X2H=P^2/(2m)+\tfrac12m\omega^2X^2. Is spatial translation generated by PP a symmetry? Evaluate the relevant commutator.
Solution

Using [X,P]=iℏ[X,P]=i\hbar,

[P,X2]=[P,X]X+X[P,X]=−2iℏX.\begin{aligned} [P,X^2] &= [P,X]X +X[P,X] \\ &=-2i\hbar X. \end{aligned}

Hence

[P,H]=12mω2[P,X2]=−iℏmω2X.\begin{aligned} [P,H] &= \frac12m\omega^2[P,X^2] \\ &= -i\hbar m\omega^2X. \end{aligned}

For ω≠0\omega\ne0 this is nonzero, so translations are not a symmetry and momentum is not conserved. The oscillator center selects a preferred position.

  1. Show that the explicitly time-dependent operator A(t)=X−tP/mA(t)=X-tP/m is a constant of motion for a free particle with H=P2/(2m)H=P^2/(2m).
Solution

The explicit derivative is

∂A∂t=−Pm.\frac{\partial A}{\partial t} =-\frac{P}{m}.

Because [P2,X]=−2iℏP[P^2,X]=-2i\hbar P,

iℏ[H,A]=iℏ[P22m,X]=Pm.\frac{i}{\hbar}[H,A] = \frac{i}{\hbar} \left[\frac{P^2}{2m},X\right] = \frac{P}{m}.

The two terms cancel:

∂A∂t+iℏ[H,A]=0.\frac{\partial A}{\partial t} + \frac{i}{\hbar}[H,A] =0.

Thus AHA_H is constant even though the Schrödinger-picture expression has explicit time dependence. This is why the commutator-only shortcut would be insufficient.

  1. Suppose Q∣i⟩=2ℏ∣i⟩Q|i\rangle=2\hbar|i\rangle, Q∣f⟩=5ℏ∣f⟩Q|f\rangle=5\hbar|f\rangle, and [Q,T]=3ℏT[Q,T]=3\hbar T. Does the QQ selection rule force ⟨f∣T∣i⟩\langle f|T|i\rangle to vanish?
Solution

The necessary condition for a nonzero matrix element is

qf−qi=qT.q_f-q_i=q_T.

Here

qf−qi=5ℏ−2ℏ=3ℏ=qT.q_f-q_i = 5\hbar-2\hbar =3\hbar =q_T.

The matrix element is symmetry allowed. The selection rule does not prove that it is nonzero; dynamics or another symmetry may still make it vanish.

  • B. C. Hall, Quantum Theory for Mathematicians, Springer, 2013.
  • M. Reed and B. Simon, Methods of Modern Mathematical Physics I: Functional Analysis, revised and enlarged ed., Academic Press, 1980.
  • J. J. Sakurai and J. Napolitano, Modern Quantum Mechanics, 3rd ed., Cambridge University Press, 2020.
  • M. H. Stone, “On one-parameter unitary groups in Hilbert space,” Annals of Mathematics 33, 643–648 (1932).