⚛️ quantum physics

From Reversible Quantum Dynamics to Statistical Probability: A dynamical solution to the origin of probability and Hilbert's sixth problem

This paper presents an exact dynamical solution to Hilbert's sixth problem by demonstrating how irreversible statistical probability and thermalization emerge from reversible unitary quantum dynamics through system-environment entanglement and the thermodynamic limit, without requiring random postulates.

Original authors: Wei-Min Zhang

Published 2026-09-17

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Wei-Min Zhang

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

For over a century, physicists have wrestled with a fundamental puzzle: how does the messy, unpredictable world of heat and chance emerge from the clean, clockwork laws of motion that govern atoms? At the smallest scales, the universe follows strict rules where cause always leads to effect, and time can theoretically run backward without breaking anything. This is the realm of quantum mechanics, where a system's state is described by a wave that evolves in a perfectly reversible way. Yet, when we look at a cup of coffee cooling down or gas spreading through a room, we see a one-way street. Things get hotter, mix, and settle into a state of equilibrium, never spontaneously un-mixing or heating up on their own. This clash between the reversible laws of the very small and the irreversible flow of time we experience has been a central mystery since the days of the great mathematician David Hilbert, who famously asked how the statistical laws of thermodynamics could possibly arise from the deterministic motion of particles.

The core of the problem lies in a question of perspective. If the entire universe is a closed system governed by reversible laws, why does a small part of it, like a single atom or a molecule, seem to lose its memory of the past and settle into a random, statistical state? For decades, the standard explanation relied on assuming that the environment surrounding a system was already a hot, chaotic soup of particles. But this approach simply pushed the problem of randomness back a step, assuming the very thing it was trying to explain. The big question remained: can true randomness and the arrow of time appear from a completely pure, ordered, and deterministic starting point, without any pre-existing chaos?

A new study by Wei-Min Zhang at National Cheng Kung University offers a definitive answer, showing exactly how this transformation happens. The researchers built a mathematical model of a system interacting with its surroundings, treating the pair as a single, closed "universe" that starts in a perfectly pure state with no randomness at all. In this model, the system and its environment are linked by forces that allow them to exchange energy. The key discovery is that even when the starting point is perfectly ordered, the act of looking only at the system while ignoring the environment creates a fundamental break in the flow of information. As the system and environment interact, they become inextricably linked, a process known as entanglement. While the total universe remains in a pure, reversible state, the part we observe loses its connection to its own past. The information about its initial state doesn't disappear; it leaks into the countless, invisible degrees of freedom of the environment, becoming inaccessible.

This leakage is what creates the illusion of randomness. The study demonstrates that if the environment has a continuous spectrum of frequencies—meaning it can absorb energy at any value rather than just specific, discrete steps—the system's memory of its starting point fades away completely over time. The system forgets where it began and settles into a stable, statistical state known as a Gibbs state. This is the familiar state of thermal equilibrium, where the system behaves as if it has a specific temperature and follows the laws of probability, even though the entire universe it belongs to is still following strict, deterministic laws. The researchers found that this transition is not a matter of chance or an added rule; it is a direct, inevitable consequence of the system's interaction with a vast, continuous environment.

Crucially, the paper rules out the idea that this process requires the environment to start as a hot, mixed-up bath. The team showed that even if the environment begins in a pure, squeezed vacuum state—a state that is perfectly ordered but possesses a specific type of quantum non-classicality—the same result occurs. The interaction itself generates the statistical weights that look like probability. However, the study also identifies a clear condition where this process fails: if the environment has gaps in its ability to absorb energy, or if the connection between the system and environment is too strong, the system can get "stuck" in a localized state. In these cases, the information never fully leaks away, the system remembers its past, and it never reaches thermal equilibrium. This provides a precise boundary for when the arrow of time emerges and when it does not.

The findings resolve a long-standing debate by showing that probability is not an extra ingredient added to the laws of physics, but rather the natural statistical structure of a part of a larger whole. When you isolate a piece of a complex, entangled system, you are left with a description that looks random because you have lost access to the rest of the puzzle. The "arrow of time" is simply the result of this information flowing out into an infinite, continuous environment, never to return. This work provides an exact, testable path from the reversible dynamics of the quantum world to the irreversible, statistical world we live in, finally connecting the deterministic laws of the atom to the thermodynamic laws of the everyday world without needing to assume randomness at the start.

Technical Summary: From Reversible Quantum Dynamics to Statistical Probability

Problem Statement
This paper addresses the core of Hilbert's sixth problem: the derivation of irreversible statistical probability and thermodynamics from the reversible, deterministic dynamics of a closed system. Historically, statistical mechanics relied on postulates such as the equilibrium assumption or the introduction of a thermal bath and molecular chaos (Stosszahlansatz) to explain irreversibility. However, these approaches face foundational challenges, including Loschmidt's reversibility paradox and Zermelo's recurrence paradox, which arise from the conflict between deterministic microscopic laws (Newtonian or Schrödinger) and macroscopic irreversibility. The central question is whether statistical probability can emerge strictly from unitary quantum evolution without imposing probabilistic interpretations or initial thermal ensembles.

Methodology
The author employs an exact open quantum system framework based on quadratic bosonic or fermionic Hamiltonians bilinearly coupled to arbitrary environments. The methodology proceeds through three distinct analytical steps:

Exact Master Equation Derivation: Starting from a closed "universe" (system + environment) evolving unitarily, the author derives the exact equation of motion for the reduced density operator ( $\rho_s$ ) by tracing out the environmental degrees of freedom. This is achieved using path integrals over coherent-state complex spaces (for bosons) and Grassmann spaces (for fermions), avoiding perturbative approximations.
Green's Function Formalism: The reduced dynamics are characterized by two non-perturbative Green's functions: a dissipative propagator $u(t, t_0)$ and a fluctuation correlation function $v(t, t)$ . These functions satisfy integro-differential equations involving memory kernels determined by the system-environment spectral density and initial state correlations.
Progressive Limit Analysis: The paper analyzes three specific scenarios to isolate the origin of probability:
- Case (i): A single system mode coupled to a finite number of environmental modes, both in pure states (system in a coherent state, environment in squeezed vacuum).
- Case (ii): A single system mode coupled to an environment with a continuous spectrum (infinite degrees of freedom), still starting from pure squeezed vacuum states.
- Case (iii): Both system and environment possess continuous spectra (macroscopic limit).

Key Results

Emergence of Mixed States from Pure States: In a finite system with a pure initial composite state, the reduced density operator evolves into a "displaced-squeezed quasi-Gibbs state." While the total entropy remains zero, the subsystem entropy becomes non-zero, representing bipartite entanglement entropy. However, in finite systems, Poincaré recurrences prevent true irreversibility.
The Role of Non-Classicality: The paper demonstrates that if the environment is in a classical state (e.g., vacuum with no squeezing), the system remains in a pure coherent state, and no thermalization occurs. Irreversibility and the generation of statistical weights require the initial environment to possess non-classical correlations (e.g., squeezing).
Thermalization via Continuum Limit: When the environment possesses a continuous spectral density without isolated bound-state poles, the dissipative propagator $u(t, t_0)$ vanishes in the long-time limit. This leads to the complete loss of information about the system's initial state. Consequently, the reduced density operator converges to a renormalized Gibbs state:
$\lim_{t \to \infty} \rho_s(t) = Z_s^{-1} \exp(-\beta_{\text{eff}} \tilde{H}_s)$
Here, $\tilde{H}_s$ is a non-perturbatively renormalized Hamiltonian, and $\beta_{\text{eff}}$ is an effective temperature determined by the steady-state fluctuation correlations.
Obstructions to Thermalization: The paper identifies that strong system-environment couplings or gapped spectral densities can generate localized bound states. In these cases, the propagator $u(t, t_0)$ retains non-decaying terms, memory of the initial state is preserved, and thermalization fails.
Generalized Fluctuation-Dissipation Theorem: The derivation naturally yields a generalized non-equilibrium fluctuation-dissipation theorem as a consequence of the unitary evolution of the closed composite, which reduces to standard forms (Callen-Welton, Green-Kubo) in equilibrium limits.

Significance and Claims
The paper claims to provide an exact, dynamical solution to the transition from determinism to statistics for quadratic quantum systems with spectral continuum requirements, addressing a core aspect of Hilbert's sixth problem. Its primary contributions are:

Derivation without Postulates: It establishes that statistical probability and the Gibbs distribution do not require an initial thermal ensemble or a probabilistic interpretation of the wavefunction. Instead, they emerge as the objective statistical structure of a subsystem when inaccessible environmental degrees of freedom are traced out.
Resolution of Paradoxes: The "time arrow" emerges from the conjunction of local reduction (tracing out the environment) and the continuum limit. In a continuum, information dissipates into infinitely many modes, preventing rephasing and recurrence, thereby resolving Loschmidt's and Zermelo's paradoxes for open subsystems within the specified model class.
Mechanism of Irreversibility: Irreversibility is identified as a result of "internal causality breaking" in the reduced dynamics, driven by non-local time correlations between forward and backward propagators when the environment is integrated out. This breaking is induced by non-classical initial correlations (e.g., squeezing) rather than classical randomness.
Testability: The author asserts that their results are testable. The required conditions (quadratic models, spectral continuum) are realizable in experimental settings such as quantum optics and solid-state transport. The theory predicts measurable quantities, including effective temperatures, renormalized frequencies, and two-time correlation functions of covariances.

The paper concludes that probability is not an additional postulate but a dynamical consequence of unitary evolution in a composite system where a part is isolated from a continuum environment, specifically within the framework of quadratic models and spectral continuum requirements. This framework locates probability within quantum dynamics and identifies irreversibility as the combined effect of exact local reduction and an infinite continuum.

Technical Summary: From Reversible Quantum Dynamics to Statistical Probability

More like this