🔢 mathematics

Consistent histories and the ultrametric on infinite tensor products

This paper constructs a complete ultrametric geometry on the space of quantum histories by demonstrating that the decay of normalized interference in the infinite registration limit is governed by the accumulated Bures distance of quantum records, thereby establishing a "consistency rate" that defines superselection sectors and quantifies the distinguishability of cumulative records.

Original authors: Andrew Lesniewski

Published 2026-09-30

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Andrew Lesniewski

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

In the quantum world, particles do not follow a single, definite path like a car on a highway. Instead, they exist in a state of many possibilities at once, a condition known as superposition. When these possibilities interact with their surroundings, they leave behind traces, or records, much like footprints in the snow. Over time, as a system evolves, these records accumulate. A central idea in modern physics, called decoherence, explains how these accumulating records cause the strange quantum possibilities to fade away, leaving behind the single, definite reality we experience every day. The process is often described as the universe "choosing" a history, but physicists have long struggled to define exactly when and how one history becomes distinct from another, especially when the differences between them are tiny and build up slowly over time.

A recent paper by Andrew Lesniewski tackles this problem by looking at the geometry of these histories. He treats the sequence of events in a quantum system not just as a list of choices, but as a path through a vast mathematical space. The core of his work is a new way to measure how far apart two different histories are from each other. In this framework, every time a system makes a choice, it leaves a record in a fresh, separate memory bank. If two histories make different choices, their records become slightly different. As time goes on, these small differences add up. Lesniewski shows that the total difference between two histories is not just a simple sum, but follows a specific, hierarchical structure known as an ultrametric. This structure acts like a ruler that measures the rate at which two histories become truly distinct, rather than just measuring the distance between them at a single moment.

The paper reveals a sharp dividing line in how these histories separate. If the accumulated differences between two histories grow slowly enough, they remain part of the same "family" of possibilities, and the quantum interference between them persists. However, if the differences grow fast enough, the two histories split into completely separate, non-interacting worlds. This is not a gradual fading but a sudden, absolute separation. The author proves that this split is governed by a specific mathematical rate. He defines a value, which he calls the consistency rate, that tells us exactly how quickly two histories become distinguishable. This rate is a single number that captures three different physical realities at once: it measures the distance between the worlds, it describes how fast the quantum interference between them vanishes, and it indicates how difficult it is to tell the two histories apart using the best possible measurement.

One of the most striking findings is that this separation can happen even when the histories are still very similar in a traditional sense. There is a "marginal" zone where two histories belong to different, separate worlds, yet the rate at which they separate is so slow that it appears almost zero. In this zone, the histories are technically distinct, but they do not separate in a way that is easily detectable by standard measures. The paper shows that the geometry of these histories is hierarchical. Just as a tree has branches that split into smaller twigs, the space of all possible histories is organized into nested layers. Histories that are very close in their long-term behavior are grouped together, while those that diverge quickly are placed in separate branches. This structure exists regardless of whether the system is finite or infinite, and it does not require the outcomes to be perfectly distinct at every single step.

The research also connects this abstract geometry to real physical systems, specifically looking at how isolated groups of particles can generate their own internal records. In some models, a large collection of particles acts as its own environment, with one part of the system recording the history of another. The paper demonstrates that the same mathematical rules apply here: the accumulation of these internal records leads to the same sharp separation of worlds. However, the author is careful to distinguish between the idealized model used for the proof and the complex reality of finite systems. In a real, finite system, there is a limit to how many distinct records can be stored, much like a hard drive has a maximum capacity. The paper shows that while a finite system can store a vast number of distinguishable histories, it cannot store an infinite number. The infinite model used in the proof serves as a perfect limit, showing what happens when the capacity for records is unlimited, turning approximate differences into exact, permanent separations.

Ultimately, the work provides a precise language for describing the emergence of classical reality from quantum mechanics. It moves beyond the idea that decoherence is just a vague process of "fuzziness" fading away. Instead, it offers a rigorous, geometric way to quantify exactly when two paths in the quantum world become two separate, non-interacting realities. The consistency rate acts as a universal gauge, telling us whether two histories are still part of the same quantum story or if they have branched off into entirely different ones. By linking the speed of record accumulation to the geometry of the space of histories, the paper reveals that the structure of our reality is encoded in the rate at which the universe keeps its memories.

Technical Summary: Consistent Histories and the Ultrametric on Infinite Tensor Products of Hilbert Spaces

Problem Statement
The consistent histories formulation of quantum mechanics assigns classical probabilities to sequences of coarse-grained events when interference terms vanish. In finite models, these interference terms (Dowker-Halliwell ratios) rarely vanish exactly; their smallness alone does not supply a well-defined asymptotic geometry on the space of histories. While recent work by Pilatowsky-Cameo et al. [16] demonstrated how decoherence and record accumulation can emerge within isolated many-body systems, a rigorous geometric framework describing the asymptotic separation of histories in the limit of infinite record accumulation was lacking. The paper addresses the need to construct a geometry for histories that accumulate quantum records, specifically characterizing the rate at which interference vanishes and how this relates to the separation of superselection sectors.

Methodology
The paper adopts an abstract, structural approach based on infinite tensor products of Hilbert spaces.

Branch Structure: A history $\Omega = (\Omega_1, \Omega_2, \dots)$ is assigned a branch vector $\Psi_{\Omega, N}$ at truncation time $N$ , where the record component factorizes as a product of unit vectors $r(\Omega_m)$ in a register Hilbert space $\mathcal{H}_{\text{reg}}$ .
Normalized Interference: The study focuses on the normalized Dowker-Halliwell ratio $d_N(\Omega, \Theta)$ , which cancels branch amplitudes and reduces to the cumulative fidelity of the records: $|d_N(\Omega, \Theta)| = \prod_{m \le N} F_m(\Omega, \Theta)$ , where $F_m$ is the per-step fidelity.
Defect and Bures Distance: The analysis utilizes the per-step defect $b_m = 1 - F_m$ , which is shown to be proportional to the squared Bures distance between the pure record states.
Infinite Product Limit: The paper applies the Kakutani-von Neumann dichotomy for infinite tensor products. It examines the convergence of the infinite product of fidelities (or the summability of defects) to determine whether two histories belong to the same weak equivalence class or orthogonal superselection sectors.
Rate Geometry: To refine the binary "same or different" sector alternative, the paper defines a "consistency rate" $\delta(\Omega, \Theta)$ based on the convergence exponent of the accumulated defect sequence. This exponent is derived from the condition $\sum m^{-p} b_m < \infty$ .

Key Contributions and Results

The Consistency Rate as an Ultrametric: The paper defines a function $\delta(\Omega, \Theta)$ representing the convergence exponent of the accumulated squared Bures distance. It proves that $\delta$ is a pseudo-ultrametric on the space of histories. After quotienting by the zero-distance relation, it becomes a complete ultrametric. This geometry requires neither compactness of the outcome space nor continuity of the record family.
Refinement of the Sector Dichotomy: The work refines the standard Kakutani-von Neumann dichotomy.
- If $\sum b_m < \infty$ , histories are weakly equivalent (same sector).
- If $\sum b_m = \infty$ , histories lie in orthogonal superselection sectors.
- Crucially, the paper identifies a "marginal stratum" where $\sum b_m = \infty$ (implying orthogonal sectors) but $\delta = 0$ . In this regime, the interference vanishes, but the separation rate is slower than any stretched exponential, demonstrating that sector separation and rate separation are distinct levels of asymptotic information.
Operational Interpretations: The consistency rate $\delta$ $δ$ is shown to have three equivalent interpretations:
1. The gauge-invariant sector separation metric $\tilde{d}$ introduced in [13].
2. The polynomial growth exponent of the negative logarithm of the normalized interference: $\delta = \limsup_{N \to \infty} \frac{\log(-\log |d_N|)}{\log N}$ .
3. The polynomial growth exponent of the negative logarithm of the minimum error probability (Helstrom bound) for discriminating between the cumulative records of two histories.
Classical-Quantum Dictionary: Using spin coherent states as a geometric model, the paper establishes a transfer principle where classical trajectory separation profiles $d_m^2 \asymp m^{-s}$ $d_{m}^{2} ≍ m^{- s}$ map directly to consistency rate strata:
- $s > 1$ : Merger (same sector).
- $s = 1$ : Marginal sector separation ( $\delta = 0$ ).
- $0 \le s < 1$ : $\delta = 1 - s$ .
- Positive long-time mean defect: Maximal rate $\delta = 1$ .
Finite vs. Infinite Record Capacity: The paper contrasts the infinite tensor product model with the finite-dimensional permutation sector of [16]. It demonstrates that while finite dimensions impose packing constraints (via Welch bounds) on near-orthogonal records, the infinite model idealizes the accumulation of records, replacing finite packing questions with the exact asymptotic sector alternative.

Significance and Claims
The paper claims to provide the "asymptotic idealization of continued record accumulation." Its primary significance lies in transforming the gradual, finite-scale suppression of interference into a sharp, exact asymptotic statement regarding superselection sectors. By introducing the consistency rate $\delta$ , the work provides a "rate geometry" that quantifies how fast histories become distinguishable, rather than just if they are.

The author emphasizes that this construction is a structural identification of the content of the infinite tensor product framework [13] within the context of consistent histories. It does not derive the dynamical mechanisms of specific models (like the chaotic kicked top in [16]) but rather supplies the mathematical structure that such mechanisms asymptotically realize. The ultrametric is presented not as an additional structure imposed on history space, but as the geometry already encoded in the accumulation of quantum records themselves. The paper explicitly separates these metric statements from the dynamical question of which separation profiles are realized by specific autonomous quantum systems, framing the results as a dictionary between classical separation profiles and quantum consistency rates.

More like this