Augmenting resonator relaxation by coupling. From coupled pendula to Rabi, Purcell, Förster, Kautsky and Partridge

An excited resonator relaxes faster when it is coupled to an absorber that accepts its energy and disposes of it. Two pendula joined by a spring, one damped, contain the whole structure: the excited pendulum's energy beats for weak absorber damping (classical Rabi oscillation), decays with the largest asymptotic rate 2g at the exceptional point kappa = 4g, and slows as 4g^2/kappa for strong damping (classical Purcell regime). The same dissipative-coupling structure appears in cavities, between molecules (Förster transfer), with many absorbers (coupling g sqrt(N)) and in photosynthesis (the Kautsky effect), under three conditions: coupling within the emitter's lifetime, disposal matched to the coupling, and a ready absorber; the resonator relaxes toward the absorber's temperature. A receiver far away meets none of these before the field's round trip, and standard physics predicts no change: whether the photon's eventual absorbers could still set the relaxation is Partridge's question, different in kind from every ordinary speedup. A time-reversal audit with the damper replaced by an explicit chain of masses confirms that all asymmetries sit in boundaries. Laws in which eventual absorption matters are excluded in a LambdaCDM future by Partridge's null unless the future absorbs everything; laws in which the local kind of ending matters remain open and testable in the laboratory (companion: Active range control of emitter-absorber coupling, doi:10.5281/zenodo.23701640).