Skip to content

Why Symmetry Matters

Symmetry matters because it turns quantum mechanics from a collection of separate solvable problems into an organized theory. A symmetry can identify conserved quantities, classify states, predict degeneracies, forbid transitions, and constrain the possible form of a Hamiltonian. The abstract algebraic object behind composable symmetry transformations is a group.

The practical question is not just “Can I solve this Hamiltonian?” It is also:

What transformations leave the physics unchanged?\text{What transformations leave the physics unchanged?}

A transformation represented by UU is a symmetry of a time-independent Hamiltonian when

UHU†=H.UHU^\dagger=H.

For a unitary UU, this is equivalent to

[U,H]=0.[U,H]=0.

This condition means that applying the transformation and then evolving gives the same physics as evolving and then applying the transformation. In other words, symmetry is a compatibility condition between transformations and dynamics.

Many important symmetries come in continuous families:

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

The Hermitian operator GG is the generator. For translations, the generator is momentum. For rotations, the generators are angular momentum components. For time translations, the generator is the Hamiltonian.

When a continuous symmetry leaves the Hamiltonian invariant,

[H,G]=0,[H,G]=0,

the generator is conserved in closed-system dynamics:

ddt⟨G⟩=0,\frac{d}{dt}\langle G\rangle=0,

assuming GG has no explicit time dependence.

If two observables commute, their eigenvalues can often be used together to label states. For angular momentum, the standard labels are

J2∣j,m⟩=ℏ2j(j+1)∣j,m⟩,J^2\lvert j,m\rangle = \hbar^2j(j+1)\lvert j,m\rangle,

and

Jz∣j,m⟩=ℏm∣j,m⟩.J_z\lvert j,m\rangle = \hbar m\lvert j,m\rangle.

The labels jj and mm are not arbitrary names. They encode how the state transforms under rotations. This is why angular momentum labels appear throughout atomic physics, molecular physics, scattering, many-body theory, and field theory.

Symmetry often organizes states into multiplets. If a Hamiltonian is rotationally invariant, states related by rotation have the same energy. This is why a central potential can have energy eigenstates arranged by angular momentum quantum numbers.

Degeneracy does not always imply a visible symmetry, and symmetry does not always force degeneracy in every representation. But when degeneracies are robust under symmetry-preserving perturbations, symmetry is usually the first explanation to test. The cautionary taxonomy is Accidental Symmetry.

A transition matrix element can vanish because the integrand is dynamically small, but it can also vanish by symmetry. For example, if a Hamiltonian has parity symmetry, its energy eigenstates can be chosen with definite parity. An operator with odd parity connects states of opposite parity and has vanishing matrix elements between states of the same parity.

The logic is more general: if states and operators transform in incompatible ways under a symmetry, a matrix element must vanish. This is the core idea behind symmetry selection rules and the Wigner–Eckart theorem.

Symmetry can be used constructively. Instead of writing the most general Hamiltonian and then solving it, write only terms allowed by the desired symmetries.

For a spin-1/21/2 degree of freedom, a general Hermitian two-level Hamiltonian can be written

H=aI+b⋅σ.H=aI+\mathbf b\cdot\boldsymbol\sigma.

If additional symmetries are imposed, some components of b\mathbf b may be forbidden. This simple observation scales upward: symmetry constraints organize effective Hamiltonians in atomic, condensed-matter, nuclear, and particle physics.

Some symmetry information is local, such as a commutator with the Hamiltonian. Some is global, such as how phases accumulate around closed loops in parameter space. Berry phase is the prototype:

γ=∮A(R)⋅dR.\gamma = \oint \mathbf A(\mathbf R)\cdot d\mathbf R.

The phase is not determined only by instantaneous energy. It depends on the geometry of the path and can reveal global structure that is invisible in a local snapshot.

  • Treating a transformation as a symmetry before checking whether it preserves the Hamiltonian.
  • Assuming every degeneracy must come from an obvious geometric symmetry.
  • Thinking conservation laws are only classical ideas.
  • Using angular momentum labels without stating the convention for J2J^2, JzJ_z, and phases.
  • Treating selection rules as approximate rather than exact consequences of stated symmetry assumptions.
  • E. P. Wigner, Group Theory and Its Application to the Quantum Mechanics of Atomic Spectra, Academic Press, 1959.
  • J. J. Sakurai and J. Napolitano, Modern Quantum Mechanics, 3rd ed., Cambridge University Press, 2020.
  • R. Shankar, Principles of Quantum Mechanics, 2nd ed., Springer, 1994.
  • H. F. Jones, Groups, Representations and Physics, 2nd ed., CRC Press, 1998.
  • M. Tinkham, Group Theory and Quantum Mechanics, Dover, 2003.
  1. Let [H,G]=0[H,G]=0 and suppose GG has no explicit time dependence. Show that ⟨G⟩\langle G\rangle is conserved under closed-system Schrödinger evolution.
Solution

The general expectation-value equation is

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

Both terms vanish under the stated assumptions, so d⟨G⟩/dt=0d\langle G\rangle/dt=0.

  1. A Hamiltonian is invariant under parity. An operator OO is odd under parity, POP−1=−OPOP^{-1}=-O. Show that ⟨ψ+∣O∣ψ+⟩=0\langle\psi_+\rvert O\lvert\psi_+\rangle=0 for a parity-even state P∣ψ+⟩=∣ψ+⟩P\lvert\psi_+\rangle=\lvert\psi_+\rangle.
Solution

Insert P−1P=IP^{-1}P=I and use parity invariance of the state:

⟨ψ+∣O∣ψ+⟩=⟨ψ+∣P−1POP−1P∣ψ+⟩=⟨ψ+∣(−O)∣ψ+⟩.\langle\psi_+|O|\psi_+\rangle = \langle\psi_+|P^{-1}POP^{-1}P|\psi_+\rangle = \langle\psi_+|(-O)|\psi_+\rangle.

Therefore the matrix element equals its negative, so it must vanish.