Information Recording Capacity as a Physical Constraint on Quantum Decoherence

Temporal Information Dynamics (TID) is a framework for quantum measurement built on a single physical principle: the rate at which a quantum system decoheres into an environmental mode is governed not only by energy coupling but by the mode’s capacity to form an irreversible, distinguishable record. The framework makes two contributions, kept deliberately distinct throughout. The first is descriptive. Within the Aharonov–Bergmann–Lebowitz (ABL) two-state formalism, a single mechanism, backward state crystallisation, gives a unified account of the standard measurement paradoxes, from the double-slit experiment and Schr¨odinger’s cat through Wheeler’s delayed choice, the quantum eraser, Wigner’s friend, EPR correlations, and the quantum Zeno effect. This is a redescription rather than a resolution of the measurement problem, and I am explicit about that limit. The second contribution is physical and testable. TID treats the environment as a structured network of receiving modes and proposes a restricted capacity-weighted routing rule: system-level outcome probabilities obey the standard Born rule, while only the routing of the released quantum among competing environmental modes is weighted by a normalised recording capacity ˜C. This restriction is what gives the framework its two protective properties. It is provably non-signalling, and it reduces exactly to standard physics for thermal environments at all temperatures, preserving stimulated emission, while leaving a sharp deviation in competitive routing between non-thermal modes. The capacity factorises as ˜C = ˜K · A into a phase-space concentration factor and a time-reversal alignment factor A ≤ 1 grounded in the ABL backward state, with concentration setting the magnitude of any deviation and alignment setting its angular structure.The central prediction is concrete. For a qubit symmetrically coupled to a thermal mode and a squeezed-vacuum mode at matched mean photon number, TID predicts a branching ratio Psq = cosh r/(1 + cosh r), that is 60.7:39.3 at r = 1 against the standard 50:50, with the full angular dependence ˜C(θ) = (1 − tanh2 r cos2 2θ)−1/2. I show by direct integration of the two-bath master equation, retaining the full anomalous squeezed-reservoir correlation M , that standard quantum optics gives 50:50 branching at every squeezing angle, while the qubit coherence over the same sweep is strongly angle-dependent. The anomalous physics is active but is structurally confined to the coherence sector and cannot reach the population branching, so the angular signature is non-absorbable into any recalibration of the reservoir parameters. The routing basis is fixed by einselection of the receiving modes (coherent states for lossy bosonic modes) rather than by the readout POVM, so the prediction survives photon-counting readout, and a quantum-non-demolition arm in which TID predicts a null result serves as a designed control. Finally, two candidate capacity functionals, phase-spaceconcentration and record distinguishability, agree that a squeezed mode out-competes a thermal one but disagree on coherent states and angular structure, so that a three-arm experiment, feasible with current superconducting circuits, discriminates standard quantum mechanics from both versions of the hypothesis.