Source-linked AI summary

Self-Locating Uncertainty and the Origin of Probability in Everettian Quantum Mechanics

Charles T. Sebens, Sean M. Carroll

arXiv:1405.7577v3quant-phgr-qc

TL;DR

The paper addresses how self-locating uncertainty should be handled in Everettian quantum mechanics and why standard treatments seem problematic. It develops a principle-based approach and concludes that there is a unique rational assignment of credences to quantum worlds.

  • Problem

    Standard treatments of self-locating uncertainty seem to generate a serious problem for reproducing the Born rule in quantum mechanics.

  • Method

    The paper considers an experimenter who has yet to observe the outcome and develops its argument using the Strong Epistemic Separability Principle.

  • Results

    The paper concludes that there is a unique rational way for an agent to assign credences to the quantum worlds they might inhabit.

  • Takeaways & Limitations

    The paper presents a justification for the Born rule in Everettian quantum mechanics.

  • Takeaways & Limitations

    The paper does not establish that true pre-measurement uncertainty exists when the wave function is known.

Abstract

from arXiv · show

A longstanding issue in attempts to understand the Everett (Many-Worlds) approach to quantum mechanics is the origin of the Born rule: why is the probability given by the square of the amplitude? Following Vaidman, we note that observers are in a position of self-locating uncertainty during the period between the branches of the wave function splitting via decoherence and the observer registering the outcome of the measurement. In this period it is tempting to regard each branch as equiprobable, but we argue that the temptation should be resisted. Applying lessons from this analysis, we demonstrate (using methods similar to those of Zurek's envariance-based derivation) that the Born rule is the uniquely rational way of apportioning credence in Everettian quantum mechanics. In doing so, we rely on a single key principle: changes purely to the environment do not affect the probabilities one ought to assign to measurement outcomes in a local subsystem. We arrive at a method for assigning probabilities in cases that involve both classical and quantum self-locating uncertainty. This method provides unique answers to quantum Sleeping Beauty problems, as well as a well-defined procedure for calculating probabilities in quantum cosmological multiverses with multiple similar observers.

1 Introduction

The paper addresses how probability can arise in deterministic Everettian quantum mechanics, arguing that self-locating uncertainty and the Epistemic Separability Principle support the Born rule rather than branch-counting. It extends this framework to classical duplication, quantum Sleeping Beauty scenarios, and large-universe quantum measurements.

  • Motivation: Everettian quantum mechanics presents a challenge for probability because its wave-function dynamics are deterministic and contain no collapse rule or hidden variables.The paper frames the Born rule’s origin as a longstanding problem for the interpretation.
  • Self-locating uncertainty: Observers can become self-locatingly uncertain about which approximately isolated branch they inhabit after decoherence, despite knowing their immediate circumstances and the universe’s state.This uncertainty concerns location among branches rather than uncertainty about the underlying universal dynamics.
  • The proposed principle: The Epistemic Separability Principle requires local predictions to be independent of environmental changes and is presented as sufficient to derive the Born rule.The paper’s simplified equal-amplitude argument uses environment modifications to constrain branch credences.
  • Main result: Given ESP, there is a unique rational assignment of credences to possible quantum worlds, and those credences are precisely the Born-rule probabilities.The paper characterizes these probabilities as subjective degrees of belief but objective in the sense that rational agents must assign them if ESP is correct.
  • The branch-counting problem: Treating quantum branches as equiprobable would extend classical indifference to identical-data observers, but the paper argues that this branch-counting recommendation is incorrect.The problem arises because observers on different branches find themselves in situations with identical data.
  • Extensions: A strengthened version of the principle also treats copies in classical self-locating uncertainty as equiprobable and addresses mixed quantum-classical cases.Applications include quantum Sleeping Beauty scenarios and quantum measurements involving multiple similar observers in very large universes.

2 Preliminaries: Many-worlds, Self-locating Uncertainty, and Branch-counting

Everettian branching produces multiple effectively non-interacting successors, yet observers face self-locating uncertainty before registering a definite outcome. The paper argues that branch-counting and Indifference give the wrong probabilities, while environmental irrelevance supports the Born rule.

  • The Many-worlds Interpretation: After measurement, each successor of Alice observes a definite outcome even though the universal state contains versions seeing different results.Many-worlds therefore explains definite individual experiences without modifying Schrödinger evolution.
  • The Many-worlds Interpretation: Decoherence makes wave-function components effectively non-interacting, allowing them to be treated as separate worlds with determinate observer records.Environmentally amplified outcome traces become approximately orthogonal, suppressing interference between branches.
  • Self-locating Uncertainty: Between decoherence and outcome registration, observers undergo self-locating uncertainty about which branch they occupy.These short-lived probabilities matter both before measurement and when updating beliefs after observation.
  • Branch-counting: Branch-counting and unrestricted Indifference can predict probabilities that differ from textbook quantum mechanics.The paper identifies environmental changes and branch proliferation as reasons not to equate branch number with rational credence.
  • Branch-counting: The proposed alternative treats environmental changes as irrelevant to local measurement probabilities and derives the Born rule as the rational assignment.The approach is intended to address both quantum and classical self-locating uncertainty.

3 The Epistemic Irrelevance of the Environment

The paper formulates the Epistemic Separability Principle: probabilities concerning an observer’s location in qualitatively identical subsystems should not change when only the external environment changes. Applied to Everettian branches, this principle supports unique probability assignments and the Born rule.

  • The Epistemic Separability Principle: ESP treats environmental facts outside the relevant subsystems as irrelevant to self-locating probabilities.Its motivation includes cases where remote events or changes deep inside the Earth do not alter the local observer’s relevant situation.
  • The Epistemic Separability Principle: ESP requires an observer’s credence about being in a particular subsystem to remain unchanged across universes that preserve the relevant subsystem states.The principle excludes additional internally identical copies of the observer outside the designated subsystem set.
  • Subsystems and Scope: The principle applies to finite collections of subsystems, which need not cover the whole universe or occur at the same time.The paper leaves the rigorous characterization of subsystems open, using physical-system constraints as rough guides.
  • Subsystems and Scope: In quantum mechanics, subsystems may arise from Hilbert-space factors or from decohered branches of a reduced density matrix.This differs fundamentally from classical subdivision into collections of particles with specified dynamical states.
  • Application to Everettian Quantum Mechanics: Applying ESP to Everettian measurement makes the probability of an outcome depend on the reduced density matrix rather than the surrounding environment.The derivation yields equiprobability in symmetric cases and extends to Born-rule assignments through branch transformations.

4 Varieties of Uncertainty

The paper distinguishes classical self-locating uncertainty from quantum branch uncertainty and develops Strong ESP to treat both within a unified framework. This framework recovers Indifference in classical duplication while yielding Born-rule probabilities for quantum branching and mixed cases.

  • Classical self-locating uncertainty: Classical self-locating uncertainty arises when an agent lacks knowledge of which qualitatively identical location or copy they occupy.The discussion considers spatial and temporal duplication, including copies separated across planets or times.
  • Strong ESP: Strong ESP requires probabilities to remain unchanged under irrelevant spatial, temporal, and environmental transformations of physically identical subsystems.The principle assumes these transformations are symmetries of the relevant dynamical laws.
  • Classical self-locating uncertainty: Strong ESP entails equiprobability among physically identical copies, thereby recovering Indifference for classical duplication cases.Pairwise swaps establish that any two copies are equiprobable, while the result is restricted to physically identical copies.
  • Quantum and mixed uncertainty: The strengthened principle also treats within-branch uncertainty with Indifference, while distinguishing it from uncertainty about which quantum branch an observer occupies.The paper emphasizes that quantum measurement can involve both kinds of uncertainty simultaneously.
  • Large-universe cosmology: For large universes, the observer-weighting recipe resolves ambiguity by making an observer’s probability depend on the amplitude of their branch rather than simply on observer counts or time-integrated population.The rule reduces to Indifference with multiple observers on one branch and to the Born rule with one observer per branch.

5 Probability in Practice

The paper addresses how Everettian agents can make decisions and confirm theories despite deterministic evolution and the absence of pre-measurement subjective uncertainty. It argues that post-measurement, pre-observation uncertainty supports Born-rule confirmation and can guide pre-measurement action, while acknowledging unresolved issues about the latter.

  • Motivation: Everettian quantum mechanics faces practical and epistemic problems concerning Born-rule-guided decisions and inference from observed long-run frequencies.The paper frames these as questions about action under future quantum measurements and theory confirmation.
  • Decision-making: The paper argues that agents can act as if a single outcome will occur with Born-rule probability because their successors will approve such choices during the post-measurement period.The argument concerns distributing goods among successors rather than making a decision under ordinary uncertainty.
  • Theory confirmation: Although pre-measurement subjective uncertainty is absent when the wave function is known, post-measurement pre-observation uncertainty is sufficient for ordinary confirmation to proceed.The authors follow Greaves in arguing that the absence of pre-measurement uncertainty does not refute Everettian quantum mechanics.
  • Decision-making: A proposed betting strategy is limited because it requires justification of preference alignment and assumes successors care narrowly about their own branches.The paper notes that broader concern for average successor wealth undermines the strategy’s straightforward appeal.
  • Theory confirmation: Observing an up result can raise Alice’s credence in ΨP↑ from 0.5 to 0.9, with repeated excess up results providing further confirmation.The paper presents this updating as evidence that Everettian probabilities can support empirical theory testing.
  • Theory confirmation: Self-locating observations can support Everettian quantum mechanics over a competing theory when that theory assigns lower probability to the observed sequence.If both theories assign the same probability, the data do not distinguish between them.

6 Comparison to Other Approaches

The paper presents an epistemic derivation of the Born rule and compares it with envariance-based and decision-theoretic approaches. It emphasizes a more explicit treatment of self-locating probabilities and a single principle applicable to both classical duplication and quantum branching.

  • The paper’s approach: The paper’s main purpose is to derive the Born rule using a purely epistemic approach to probabilities in Everettian quantum mechanics.It positions this project against concerns about existing derivations rather than claiming their mathematical methods are wholly unrelated.
  • Comparison with envariance: Like Zurek’s argument, the derivation uses environment transformations and entanglement to recover Born-rule probabilities for ordinary measurement scenarios.The relevant swap leaves subsystem and environment states unchanged while altering their entanglement, yielding equal probabilities in the symmetric case.
  • Comparison with envariance: The authors argue that Zurek’s treatment does not explain how probabilities arise in a deterministic Everettian theory or clearly identify what the probabilities concern.They attribute this gap to the absence of a discussion of self-locating uncertainty.
  • The paper’s approach: The paper claims to provide a more thorough justification of its assumption and a more philosophically careful treatment of the probabilities involved.Its assumption, ESP-QM, is described as similar to Zurek’s two assumptions taken together.
  • Comparison with decision theory: Decision-theoretic approaches rely on assumptions such as Measurement Neutrality or Equivalence, which the authors argue conflict with prima facie reasonable Indifference or beg the question against branch counting.The authors instead reject Indifference because it conflicts with ESP, a single general epistemic principle.
  • Comparison with decision theory: The paper’s approach is presented as applicable to both classical duplication and quantum branching while remaining consistent with Indifference in classical self-locating cases.Strong ESP is said to explain why the two kinds of cases are handled differently.

7 Conclusion

The paper justifies the Born rule through self-locating uncertainty and Strong ESP, unifying classical Indifference with quantum probabilities. It also identifies several areas requiring further philosophical and theoretical development.

  • 7 Conclusion: Indifference is reasonable in classical self-locating scenarios but yields strange recommendations in quantum contexts, motivating a different principle.The paper contrasts classical equiprobability with the quantum treatment supplied by Strong ESP.
  • 7 Conclusion: Strong ESP extends classical Indifference to quantum measurement, yielding the Born rule in the many-worlds interpretation.The principle is presented as physically transparent and as a bridge between classical and quantum uncertainty.
  • 7 Conclusion: Probability in a deterministic theory is explained by evolution from perfect knowledge to unavoidable self-locating uncertainty.The account treats self-locating uncertainty as fundamental to the emergence of probabilities.
  • 7 Conclusion: The approach provides a unified perspective on uncertainties in classical and quantum contexts, with implications for large-universe cosmology and quantum-mechanical foundations.The paper applies this perspective to cases involving both classical and quantum self-locating uncertainty.
  • 7 Conclusion: Future work should further justify and precisely formulate Strong ESP, defend reduced density matrices as subsystem representations, clarify pre- and post-measurement decision making, and analyze branching assumptions.The conclusion also notes the need for a more detailed account of when and how branching occurs.

A What’s Not Wrong with Branch-counting

The section argues that diachronic inconsistency and Dutch-book objections do not establish a distinctively quantum problem for Indifference. Branch-counting remains reasonable in the relevant classical and quantum cases.

  • Branch-counting and diachronic change: Indifference produces changing credences across successive measurements because later observers face more indistinguishable branches.After the second measurement, Alice regards herself as twice as likely to be on an up branch because there are twice as many up branches.
  • Branch-counting and diachronic change: These credence shifts can make deliberative action difficult in a richly branching Everettian multiverse.The section connects frequent branching with near-impossible deliberative action.
  • Dutch-book argument: Indifference can generate an Everettian Dutch book in which individually fair bets guarantee a loss on every branch.The example offers separate bets at different times that Alice judges fair, although accepting both produces a $5 loss on each branch.
  • Dutch-book argument: The authors argue that this Dutch book does not uniquely undermine Indifference because analogous diachronic Dutch books arise in classical duplication cases.The defender of Indifference may reject the inference from such a Dutch book to irrationality.
  • Classical comparison: The charge of diachronic inconsistency therefore does not raise a distinctively quantum problem for Indifference.The comparison with classical fission supports treating the objection as non-specific to quantum branching.

B Circularity

The section examines whether deriving the Born rule from reduced density matrices is circular. It presents an alternative subsystem-based justification while acknowledging unresolved questions about branching and representation.

  • The circularity objection: A central objection is that using reduced density matrices may presuppose the Born rule when deriving it.The concern targets the physical interpretation of reduced density matrices and their role in representing subsystems.
  • Remaining concerns: The derivation uses reduced density matrices to represent subsystem states, but the authors acknowledge that this identification can be doubted.The physical justification offered is that the reduced density operator gives the correct measurement statistics for subsystem measurements.
  • Alternative justification: The authors argue that reduced density matrices are mathematically well-defined within quantum theory and need not be justified through probabilities or measurements.They propose an alternative justification that does not rely on probabilistic interpretation.
  • Alternative justification: The proposed subsystem representation should combine with other subsystem states and connection facts to determine the total state, while determining isolated subsystem evolution.For an isolated subsystem, the total evolution factorizes and the environment operator becomes irrelevant to the subsystem’s reduced state evolution.
  • Remaining concerns: The paper does not resolve how wave-function patterns become distinct worlds or observers who can be uncertain about which world they occupy.It treats the existence of such branching structure as an assumption and identifies its explanation as primarily metaphysical.

C Generalization of the Born Rule Derivation

The section generalizes the Born-rule derivation from simple cases to arbitrary rational squared-amplitudes and then extends it to irrational cases by continuity. The proof uses environment transformations that preserve the local subsystem state.

  • Equal-amplitude construction: Equal-amplitude components are shown to be equiprobable, so counting components associated with each detector state yields the Born rule.The method extends the earlier arguments to N orthogonal branches with arbitrary rational squared-amplitudes.
  • Irrational amplitudes: The proof initially covers rational squared-amplitudes, then extends to irrational cases by assuming probabilities vary continuously with small amplitude changes.Rational squared-amplitudes are sufficient because rational values are dense among the reals.
  • General state: The generalized derivation assigns each detector state d_k a probability equal to its amplitude squared, c_k^2.When multiple branches share the same detector state, their amplitude-squared contributions are summed.
  • Equal-amplitude construction: Environment transformations convert unequal-amplitude branches into equal-amplitude components while leaving the Alice+Detector reduced density matrix unchanged.This preserves the local subsystem information relevant to the probability assignment.
  • Conclusion of the proof: The proof concludes that probabilities for the different detector states in the general state are given by the Born rule.The transformed and original states agree on the Alice+Detector reduced density matrix, supporting the transfer of probabilities.
Loading 1405.7577v3…