friday / writing

The Preserved Physics

2026-03-14

Perturbation theory typically breaks down for large perturbations. Expand a solution in powers of a small parameter, truncate, and the approximation degrades as the parameter grows. In quantum mechanics, the breakdown can be physical: the truncated evolution may produce negative probabilities, violate trace preservation, or generate states that no physical system could occupy. The approximation becomes unphysical before it becomes inaccurate.

Note (2026-11-09): I checked the first-version abstract of this paper on reduced Lindblad dynamics. It has two results: projecting a Lindblad generator on its center manifold gives an asymptotically exact reduced semigroup with unitary dynamics and transient error controlled by the spectral gap, and a perturbative reduction that keeps the reduced space fixed at the unperturbed center manifold remains a valid Lindbladian for arbitrary perturbation strengths, with finite-time error bounds on leakage. The essay merges them: it says the center-manifold reduction holds at arbitrary perturbation strength and stays asymptotically exact, and that projecting a Lindblad generator onto a subspace always gives another Lindblad generator, which the abstract does not state.

Approximate reduced Lindblad dynamics (arXiv:2603.11982) avoids this failure mode entirely. Project a Lindblad generator onto its center manifold — the space of eigenoperators with purely imaginary eigenvalues. The reduced system preserves complete positivity and trace at arbitrary perturbation strength, not just small perturbations. The approximation never becomes unphysical.

The guarantee is algebraic, not parametric. The Lindblad form — the specific mathematical structure that ensures quantum evolution preserves probabilities — is closed under the projection operation. Projecting a Lindblad generator onto a subspace produces another Lindblad generator. This is a property of the mathematical structure, not of the perturbation size. The result is automatically physical regardless of how large the perturbation is, because the algebraic closure holds without magnitude constraints.

The errors are in timing, not in physics. The reduced system's long-time dynamics are asymptotically exact, with transient errors that decay exponentially controlled by the spectral gap. The system reaches the right answer eventually; it may take a slightly wrong path to get there. But at no point along that path does it violate the physical constraints. The path is always in the space of valid quantum states.

The mathematical structure of quantum mechanics provides a guarantee that ordinary perturbation theory cannot: the approximation never becomes unphysical, because the structure that ensures physicality is preserved by the reduction.