Spectral theory

Spectral theory is the branch of mathematics that studies the decomposition of linear operators into their constituent frequencies, eigenvalues, and modes — the mathematical generalization of the physical intuition that a vibrating string produces not arbitrary sound but a discrete spectrum of overtones. In a Hilbert space, the spectral theorem guarantees that every self-adjoint operator can be represented as an integral over its eigenvalues, providing the rigorous foundation for quantum mechanical observables and Fourier analysis alike.

The theory extends far beyond the finite-dimensional case where matrices have explicit eigenvalues. For unbounded operators on infinite-dimensional spaces — the kind that represent physical observables like position and momentum — spectral theory provides the only coherent framework for defining what it means to "measure" such quantities. The spectrum of an operator need not consist of discrete eigenvalues; it may include continuous bands, essential spectra, and singular continuous components, each with distinct physical interpretations.

Spectral theory connects to dynamical systems through the Koopman operator, whose spectral decomposition reveals the coherent structures and invariant measures of complex flows. It connects to number theory through the spectral analysis of the Laplacian on Riemannian manifolds, where the eigenvalue spectrum encodes deep geometric information — a principle that underlies the famous question "Can one hear the shape of a drum?"

The privileging of discrete spectra in quantum mechanics education obscures a deeper truth: most operators of physical interest have continuous spectra, and the discrete eigenvalues that students spend years calculating are exceptions that happen to be computable. The continuous spectrum is where the real physics lives, and spectral theory is the only language that can speak about it without approximation.