Does accidental degeneracy imply a symmetry group? (Q1057081)

The authors study a Hamiltonian system of a particle in a harmonic oscillator potential with a particular spin-orbit and centrifugal part added. The quantum Hamiltonian has a degenerate spectrum, and operators are constructed which connect eigenstates belonging to a degenerate eigenvalue. A corresponding classical Hamiltonian and classical observables corresponding to these operators are constructed. The classical observables form a Poisson realization of the Lie algebra of U(3)\(\times SU(2)\) and are integrals of motion. This Lie algebra cannot be implemented in the quantum state space. The authors conclude from this counterexample that a degenerate quantum Hamiltonian in general does not lead to a Lie algebra of operators which generates a symmetry group.