Can Quantum Interference Arise from the Dynamics of Time? A Matter-Free Local Time-Rate Field Study

Can two-path quantum-interference phenomenology be understood as a property of a local time-rate field before an individual result is fixed? We investigate this question in a matter-free effective dynamical branch of the Enchan framework. A real local time-rate field, initialized with field and velocity data, propagates through two apertures and produces a non-additive, oscillatory screen-energy-flux pattern without primitive particles, a primitive wavefunction, Born weights, or a collapse operation. Its normalized cross term is 0.533636, 0.529264, and 0.526779 on three successively refined grids. A linear-response control retains the effect, identifying a broader coherent-field mechanism. We then evolve a coupled field/path-tag model in which the passing field writes opposite tag orientations near the two slits. Across a frozen five-strength sweep, the independently measured route-tag overlap and unresolved interference decrease together; their Pearson correlations are 0.996786 and 0.998675 on two numerical levels. The screen is computed from the evolved fields without multiplying a reference fringe by a tag-overlap factor. The additional tag channel and slit-local antisymmetric coupling are explicit minimal model assumptions, not a unique derivation from the canonical scalar ontology. The physical comparisons are route removal, path tagging with both routes open, and localized screen recording. A half-amplitude control reduces flux to 24.64% while retaining the normalized cross term. A third calculation reuses the exact Q0 field and defines a nonnegative screen event measure from the accumulated positive forward energy flux, J_+(y) = integral max(F_E,x, 0) dt. One localized screen coordinate is drawn per independent preparation from the normalized $J_+$ distribution. In a fixed 600-event realization, the binned hit histogram has Pearson correlation 0.9553 and total-variation distance 0.0949 relative to the binned event measure. This finite-sample agreement is a consistency check of the explicit forward-flux readout, not an independent validation and not a derivation of the Born rule. Interference, path-tagging, and one-hit event realization are therefore supported at their separately stated model levels, while microscopic detector selection and experimental confirmation of the proposed ontology are outside the scope of this study. The release includes the manuscript PDF, a standalone reproducibility ZIP with manuscript sources, frozen numerical solvers and inputs, lightweight results and figures, an English reproduction manual, LICENSE, and SHA-256 checksums. The manuscript is dated 18 September 2026. No private repository, external dataset, or GPU is required. Enchan Research & Verification License v1.0. Copyright (c) 2026 Mitsuhiro Kobayashi. All rights reserved.