Is the Reduced Density Matrix Complete for Local Evolution? A Relational-Context Test Beyond Standard Nonlinear Quantum Mechanics

Overview This work introduces and analyzes an operational test of a foundational assumption frequently made in quantum theory and quantum device modeling: that the reduced density matrix of a subsystem is a sufficient state variable for predicting its future reduced evolution under fixed, declared controls. We formalize this assumption as Local Evolution Completeness (LEC) and show how it can be tested experimentally. The central question is not whether quantum mechanics is linear, nor whether completely positive trace-preserving (CPTP) maps exist, nor whether global evolution is unitary. Instead, the question is more subtle and operational: If two globally distinct preparations share the same reduced density matrix at an initial time, must they necessarily yield identical reduced dynamics under identical device controls? This question probes the predictive sufficiency of reduced states rather than the linearity of evolution. It isolates a distinct experimental axis from traditional nonlinear quantum mechanics tests. Core Concept: Local Evolution Completeness (LEC) We define Local Evolution Completeness (LEC) as follows: If two global states produce the same reduced density matrix ρS (0) for a subsystem S, then under fixed device dynamics they must produce identical reduced states ρS (t) at all future times. In standard modeling practice, this assumption is often implicit. One typically writes: ρS (t) = Φₜ (ρS (0) ), where Φₜ is a CPTP map. This implicitly assumes that the reduced density matrix at time zero is a sufficient statistic for future local evolution. The manuscript demonstrates that this assumption is not universally valid. Strict-Local Null Result As a baseline, we prove a strict-locality theorem: If the applied Hamiltonian factorizes as H = HS ⊗ IE, then LEC holds exactly, regardless of correlations with ancillary degrees of freedom. This provides a strong experimental null test. Any deviation under strictly local control must arise from uncontrolled couplings or miscalibration. Explicit Process-Tensor Counterexample The paper provides a fully explicit two-time process-tensor construction in which: The global evolution is strictly unitary. All intermediate maps are completely positive and trace-preserving. Two distinct intervention histories produce identical reduced states at time t₀. Yet the reduced states at time t₁ differ. This demonstrates that: There exist processes in standard quantum mechanics for which no reduced map Φ exists that depends only on ρS (0). Thus, CPTP slices may exist while LEC fails. This reframes the issue in the modern language of process tensors: the reduced state at a single time is not, in general, a sufficient statistic for multi-time prediction. Solvable Qubit–Ancilla Model To make the discussion concrete and experimentally tractable, the paper develops a fully solvable qubit–ancilla model: Two global states are constructed with identical reduced marginals. Exact reduced trajectories are derived analytically. Bloch-sphere expressions are provided in closed form. Trace-distance witnesses are calculated explicitly. This provides a transparent and minimal demonstration of LEC failure in finite dimension. Microscopic Realizations of Embedding Dependence The manuscript presents multiple microscopic realizations of the embedding variable responsible for LEC failure: Initial-correlation invariant modelTwo global states share identical reduced marginals yet differ in correlation invariants. Under fixed global unitary dynamics, these correlations generate distinct reduced trajectories. Control-reference modelThe embedding variable is realized as a phase of a drive oscillator. Even when declared controls are identical, the effective reduced generator depends on relational reference data not encoded in ρS (0). These constructions show that LEC failure can arise without modifying the Schrödinger equation, without invoking nonlinear dynamics, and without abandoning global unitarity. Experimental Protocol The manuscript provides a detailed experimental framework: Preparation of globally distinct states with identical reduced marginals. Strict-local sanity test to validate baseline behavior. Time-resolved Bloch vector tomography. Likelihood-ratio statistical testing. Fisher-information scaling analysis. Explicit bounds on detectable deviations. The proposal is implementable in current qubit platforms such as superconducting circuits or trapped-ion systems. Exclusion of Ordinary Explanations The work systematically addresses and excludes standard objections: Uncontrolled system–environment coupling. Calibration drift. Initial-correlation artifacts. Convex-linearity violations. Misinterpretation as Weinberg-type nonlinear quantum mechanics. A strengthened SE-closure hypothesis is introduced to clarify the operational meaning of the test. Conceptual Significance This study does not propose a modification of quantum mechanics. Instead, it isolates and operationalizes a structural question: Is the reduced density matrix a sufficient state variable for future local prediction, or merely a compression of richer multi-time structure? In process-tensor language, LEC corresponds to a compression assumption: that the operational slice of the process can be reduced to a single CPTP map acting on ρS (0). The manuscript shows explicitly that this compression can fail even under globally unitary dynamics. What This Work Does Not Claim It does not introduce nonlinear Schrödinger dynamics. It does not contradict Stinespring dilation. It does not refute CPTP structure. It does not claim violation of global unitarity. It does not assert new fundamental physics. Instead, it proposes a test of predictive sufficiency under controlled settings. Summary This work: Defines Local Evolution Completeness (LEC) precisely. Proves a strict-locality null theorem. Constructs explicit finite-dimensional counterexamples. Embeds the problem within process-tensor formalism. Provides a solvable qubit model. Proposes an experimentally realizable protocol. Supplies full statistical methodology. The result is a coherent operational framework for probing when reduced density matrices are sufficient statistics for quantum dynamics and when they are not.