⚛️ quantum physics

On the significance of Wigner's Friend in contexts beyond quantum foundations

This paper challenges the view that the Wigner's Friend paradox is unique to quantum theory by demonstrating that its core logical structure arises from a general "Restriction A"—the inability of a physical theory to provide probabilistic descriptions of all agents' observations—which also underpins broader enigmas in physics and philosophy, such as the Boltzmann brain problem.

Original authors: Caroline L. Jones, Markus P. Mueller

Published 2026-06-23

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Caroline L. Jones, Markus P. Mueller

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

The Big Idea: It's Not Just About Quantum Magic

For decades, physicists have been fascinated by a thought experiment called "Wigner's Friend." The story goes like this: A friend is inside a locked lab looking at a quantum particle. To the friend, the particle has a definite result (like "Heads"). But to Wigner, standing outside the lab, the friend and the particle are in a fuzzy, mixed-up state (a "superposition") until Wigner opens the door.

The usual conclusion is that this proves quantum mechanics is weird and that "facts" might depend on who is looking at them.

This paper argues that the weirdness isn't unique to quantum physics. The authors, Caroline Jones and Markus Müller, claim that the core puzzle of Wigner's Friend can be recreated using classical physics (the physics of everyday objects) if you introduce one specific trick: copying people.

They argue that the real problem isn't "spooky quantum entanglement"; it's a fundamental limitation in how we describe reality when multiple observers exist, especially when those observers get duplicated.

The Core Concept: "Restriction A"

The authors introduce a new term called "Restriction A."

Imagine you are trying to write a single, master "scorecard" (a joint probability distribution) that lists the results of every player in a game.

No Restriction A: You can write one scorecard that perfectly predicts what Player A sees, what Player B sees, and how their views relate to each other.
Restriction A: The rules of the game make it impossible to write that single master scorecard. The theory simply cannot give you a combined picture of what everyone sees.

The paper claims this happens in two places:

Quantum Physics: In Wigner's Friend scenarios.
Classical Physics: In scenarios where people are duplicated (cloned).

The Classical Analogy: The "Fission" Experiment

To prove their point, the authors imagine a world where humans can be perfectly copied, like a cell dividing.

The Thought Experiment:

Freya enters a machine.
The machine splits her into two identical copies: Freya-Blue (who wakes up in a blue room) and Freya-Green (who wakes up in a green room).
Both copies remember being Freya before the split.
Later, the copies are merged back into one person, who wakes up with no memory of which room they were in.

The Problem:
Before the split, Freya asks: "What will I see? Will I see blue or green?"

The Classical Physics Answer: Classical physics can describe the machine and the rooms perfectly. But it cannot tell Freya the probability of seeing blue vs. green.
Why? From the outside, there are now two Freyas. From the inside, Freya doesn't know which one she will be. There is no "master scorecard" that links the single "pre-split Freya" to the "post-split Blue Freya" and "post-split Green Freya" in a way that classical math can calculate.

The authors say this is the same structural problem as Wigner's Friend. The "Restriction A" is that the theory (classical physics) cannot give a joint description of the observations of all agents (the original Freya and the two copies).

The "Sleeping Beauty" Twist

The paper also uses a famous puzzle called the Sleeping Beauty Problem to show that even if you do try to guess the odds, different observers will disagree on the math.

The Setup: Freya and her friend Wigner are put to sleep. A coin is flipped.
- If Heads: Freya is copied into 1,000 versions. Wigner stays as 1 version.
- If Tails: Freya stays as 1 version. Wigner stays as 1 version.
The Wager: They are woken up and asked to bet on the coin flip.
The Conflict:
- Wigner (who sees the same number of copies regardless of the coin) thinks the odds are 50/50.
- Freya (who knows that if it's Heads, there are 1,000 of her to wake up) thinks it's much more likely to be Heads. If she wakes up, she is statistically more likely to be one of the 1,000 "Heads" copies than the single "Tails" copy.

The Result: They have the same information, but they calculate different probabilities. If you try to force them into a single "joint probability" (a master scorecard where they both agree on the math), the math breaks. This is Restriction A in action.

Why This Matters (According to the Paper)

The authors are challenging the idea that Wigner's Friend is a "Quantum Only" mystery.

It's About Identity, Not Just Physics: The puzzle arises because our language and logic struggle with "Who am I?" when there are multiple copies of "me." This happens in classical cloning just as much as in quantum superposition.
The "Black Box" Problem: The core issue is that some experiments create an "Epistemic Horizon" (a knowledge barrier).
- In the quantum version, the "Friend" is in a closed lab, and Wigner can't see inside without destroying the quantum state.
- In the classical version, the "Friend" is duplicated, and there is no external way to define which copy is "the" observer.
- In both cases, the theory cannot give a single, unified description of what everyone sees.

The Boltzmann Brain Connection

The paper even connects this to Cosmology and the Boltzmann Brain problem.

The Problem: In an infinite universe, it's statistically possible for a brain to randomly pop into existence with false memories of a whole life (a "Boltzmann Brain").
The Connection: If you are a Boltzmann Brain, you can't ask "Am I a real person or a random fluctuation?" from the outside. The theory cannot give you a joint probability of "Me being real" vs. "Me being a fluctuation." This is another example of Restriction A.

Summary in a Nutshell

The paper argues that Wigner's Friend is not a sign that quantum physics is broken or magical. Instead, it reveals a deep, structural limit in any physical theory (quantum or classical) when you try to describe the world from the perspective of multiple agents who might be duplicated or have conflicting views.

The Takeaway:
We often think the universe is like a movie where everyone sees the same scene. This paper suggests that in certain scenarios (like cloning or quantum labs), the universe is more like a choose-your-own-adventure book. The "author" (the physical theory) can tell you the rules of the story, but it cannot always write a single summary that explains what every character sees at the same time. This limitation is called Restriction A, and it applies to the quantum world, the classical world, and even the cosmos.

Technical Summary: On the significance of Wigner's Friend in contexts beyond quantum foundations

Problem Statement
Recent interest in the Wigner's Friend (WF) paradox, particularly through Extended Wigner's Friend (EWF) scenarios (e.g., Frauchiger-Renner, Bong et al.), has largely framed the phenomenon as a unique feature of quantum theory. These scenarios are typically interpreted as demonstrating a conflict between quantum mechanics and the "absoluteness of observed events" (AOE), suggesting that quantum theory is incompatible with the idea that different agents can share a single, objective reality. The prevailing narrative posits that the paradox arises from uniquely quantum features such as superposition, entanglement, or the violation of Bell-type inequalities.

The authors challenge this view, asking whether the structural difficulties inherent in EWF scenarios—specifically the inability to assign a joint probabilistic description to the observations of multiple agents—are unique to quantum theory or if they arise in broader physical and philosophical contexts, including classical physics.

Methodology
The paper employs a comparative analysis of quantum thought experiments and classical analogues involving agent duplication and branching. The methodology proceeds in three stages:

Reconstruction of Quantum Scenarios: The authors review the "Local Friendliness" (LF) no-go theorem and the Frauchiger-Renner (FR) thought experiment. They analyze the metaphysical assumptions required to derive contradictions, such as AOE, Locality, No Superdeterminism, and the consistency of agents' beliefs.
Construction of Classical Analogues: The authors propose two classical thought experiments involving the duplication of agents:
- Thought Experiment 1 (Duplication and Merging): An agent (Freya) is duplicated into two copies (in blue and green rooms) and subsequently merged back into a single agent with erased memories.
- Thought Experiment 2 (The Sleeping Beauty Variant): Agents (Freya and Wigner) are multiplied into $N$ copies. In each lab, a coin is tossed; if Heads, the Freya copy is duplicated again, while the Wigner copy remains singular. Agents must assign credences to the coin outcome.
Structural Abstraction (Restriction A): The authors abstract the core difficulty of these scenarios into a formal property they term "Restriction A." They then demonstrate how this property manifests in the quantum scenarios, the classical analogues, and the cosmological Boltzmann brain problem.

Key Contributions

Definition of Restriction A: The paper introduces "Restriction A" as a structural property of physical theories. It is defined as the situation where a physical theory (even with a complete description of the experiment) fails to provide a joint probabilistic description of the observations of all agents involved.
- For $N=1$ agent: The theory cannot provide a probabilistic prediction for the agent's future private observations.
- For $N \ge 2$ agents: The theory cannot provide a joint distribution from which individual agents' observations can be derived as marginals.
Classical Reproduction of EWF Features: The authors demonstrate that the metaphysical dilemmas of EWF scenarios (specifically the failure of AOE and the inconsistency of agents' beliefs) can be reproduced in classical settings without quantum mechanics.
- In Thought Experiment 1, classical physics cannot define a random variable for "Freya's observation" from an external perspective because the referent of "Freya" becomes ambiguous after duplication. Consequently, classical physics cannot assign a probability to Freya's private experience, constituting an instance of Restriction A for $N=1$ .
- In Thought Experiment 2, the authors show that even if agents adopt Elga's Principle of Indifference to assign individual probabilities (Freya assigns $P(T) = 1/(M+1)$ , Wigner assigns $P(T) = 1/2$ ), these assignments cannot be embedded into a single joint probability distribution. This leads to a "classical no-go theorem" (Theorem 1) showing the incompatibility of Elga's Principle, the "No Update on Irrelevant Information" (NUI) assumption, and "Probabilistic Consistency" (CP).
Resource Analysis: The authors argue that these classical scenarios are not merely philosophical abstractions but are potentially implementable. They compare the resource costs of simulating these scenarios on classical computers versus quantum computers, noting that classical simulations of duplicated agents are currently more feasible than the quantum implementations required for EWF tests.
Generalization to Cosmology: The authors extend the analysis to the Boltzmann brain problem, arguing that the question "Am I a Boltzmann brain?" represents an instance of Restriction A. The answer is private to the observer but cannot be intersubjectively verified or described by a joint probability distribution within the theory.

Results

Non-Uniqueness of the Paradox: The core structural difficulty of the Wigner's Friend paradox—the inability to define a joint probability distribution for all agents' observations—is not unique to quantum theory. It arises in classical scenarios involving agent duplication where personal identity becomes ambiguous.
The "Epistemic Horizon": Restriction A applies specifically in situations featuring an "epistemic horizon," where the observations of agents cannot be jointly assessed by an external observer. In quantum EWF scenarios, this is enforced by unitary reversal of measurements; in classical scenarios, it is enforced by the ambiguity of personal identity in branching scenarios.
Incompatibility of Assumptions: The paper proves that in the classical Thought Experiment 2, the assumptions of Elga's Principle of Indifference, NUI, and Probabilistic Consistency (CP) cannot all hold simultaneously. This mirrors the contradictions found in quantum EWF scenarios but derives them from classical probability theory and self-locating uncertainty.
Distinction from Bell Violations: The authors clarify that while classical analogues cannot reproduce the violation of Bell inequalities (which is a specific quantum feature), they can reproduce the failure of the "Absoluteness of Observed Events" and the inability to form joint probabilistic descriptions.

Significance and Claims
The paper claims that the Wigner's Friend paradox should be studied not merely as a quantum foundational problem, but as a broader frontier in physics and philosophy concerning private but not intersubjectively verifiable predictions.

Broadening the Scope: The authors argue that Restriction A is a fundamental limitation of physical theories when applied to scenarios involving multiple agents with diverging perspectives (due to branching, duplication, or cosmological uncertainty).
Re-evaluating Quantum Uniqueness: While quantum theory exhibits specific forms of correlation (violating Bell inequalities) that are non-classical, the structural difficulty of assigning joint probabilities to agents in branching scenarios is a more general feature. The "quantumness" of Wigner's Friend lies in the specific implementation (e.g., unitary reversibility), but the metaphysical consequence (the failure of AOE) is a general feature of theories dealing with agent duplication.
Implications for Interpretation: The analysis suggests that resolving the Wigner's Friend paradox may require addressing the general problem of how to define "agents" and their observations in theories where personal identity is not a fixed, single-valued variable. The paper concludes that Wigner's Friend shares a common structural core with other puzzles like the Boltzmann brain problem, united by the presence of Restriction A.

The authors remain modest, noting that they do not claim classical scenarios reproduce the phenomenology of quantum interference or Bell violations, but rather that they reproduce the structural and metaphysical difficulties regarding the absoluteness of observations and the consistency of agent reasoning.