Well-definedness and robustness of Kent's late-time-measurement beables beyond toy models

Kent's Lorentzian quantum reality proposal (Phil. Trans. R. Soc. A 373, 20140241 (2015)) supplements standard unitary quantum theory with beables — definite mass-energy values — defined at each spacetime point by conditioning on the outcomes of hypothetical measurements made in the arbitrarily far future, outside that point's future light cone, with probabilities taken in an assumed asymptotic limit, thereby giving a one-world, Lorentz-covariant realist description without modifying quantum dynamics. The paper's own discussion flags unresolved technical questions and suggested directions: the existence of the asymptotic limit $\mathrm{Prob}_{\infty}(C_{S_{1}})=\lim_{S\to\infty}\mathrm{Prob}_{S}(C_{S_{1}})$ — and, in the predecessor paper's definition of the beables themselves, the independence of the limit on the chosen sequence of hypersurfaces — is assumed, not proven; realistic 3+1-dimensional models with wavepacket spreading, recoil, imperfect scattering, soft photons and nonzero massive-subsystem Hamiltonians have not been analyzed; the class of spacetimes in which massless records escape to future infinity (so that the prescription yields quasiclassical beables) is not delineated; and the natural generalizations the paper suggests — reweighted beable probabilities and coupling to gravity — are left for future work, with their consistency unexamined.

Background

In John Bell's formulation, the quantum reality problem asks what the 'beables' of quantum theory are — which physical quantities are assigned definite values, with what probabilities, at which spacetime points — so as to give a single-world, observer-independent description of reality. A 'Lorentzian' solution must do this without introducing a preferred slicing of spacetime and without conflicting with Bell-type no-go theorems, which is notoriously hard.

Adrian Kent proposed such a framework (Phys. Rev. A 90, 012107 (2014); Phil. Trans. R. Soc. A 373, 20140241 (2015)). The idea is to supplement standard unitary quantum theory with beables defined by hypothetical measurements made in the arbitrarily far future. One postulates a (fictional) late-time measurement of mass-energy density — formally of the stress-energy tensor $T_{\mu\nu}$, the operator-valued field whose components encode energy and momentum density — on a spacelike hypersurface $S$ far in the future. Writing $t_S(x)$ for the outcome, the beables (conditional expectation values of $T_{\mu\nu}$ at intermediate points $y$) are computed from the Born-rule joint distribution of $t_S(x)$ and $T_{\mu\nu}(y)$, conditioning only on the outcomes $t_S(x)$ at points $x$ outside the future light cone of $y$. Because these data lie outside $y$'s causal future, the construction is Lorentz covariant and involves no causal paradox; 'quasiclassical' systems (systems whose coarse-grained mass densities follow approximately classical equations of motion, with stochastic corrections) are expected to leave records of their location in photons and other massless particles that escape to future infinity. The full prescription takes the limit of an infinitely receding hypersurface: assuming that the limit $\mathrm{Prob}_{\infty}(C_{S_{1}})=\lim_{S\to\infty}\mathrm{Prob}_{S}(C_{S_{1}})$ exists, for each coarse-grained set $C_{S_{1}}$ of stress-energy configurations between an initial surface $S_{0}$ and a fixed later surface $S_{1}$, this defines the probability that reality between $S_{0}$ and $S_{1}$ is described by a mass distribution in $C_{S_{1}}$.

Kent illustrated the postulates in toy models set in one space and one time dimension: photons treated as pointlike particles moving at light speed, perfectly and instantaneously reflected by pointlike massive objects whose Schrödinger dynamics is switched off ($H=0$ for massive subsystems). Even these models exhibit artifacts — for instance a discontinuity in the beable description precisely along the lightlike separation between two components of a superposition. Kent's discussion flags the program's open ends: the construction presumes stress-energy measurements, Tomonaga–Schwinger evolution (the definition of unitary state evolution on arbitrary spacelike hypersurfaces) and asymptotic limits, none of which are established rigorously in realistic interacting quantum field theories in 3+1 dimensions, a gap Kent ties explicitly to the general unsolved problem of rigorously constructing such field theories; the $\mathrm{Prob}_{\infty}$ limit is assumed, not proven; realistic 3+1-dimensional models with spreading wavepackets, momentum recoil, imperfect scattering, many photons and nonzero massive-subsystem Hamiltonians are needed to test the qualitative conclusions; the prescription is contingent on cosmology (in a recollapsing 'big crunch' universe, massless records cannot escape to future infinity, so it fails); and the paper suggests natural generalizations — reweighting the beable probability assignments, or coupling the beables (which take the form of a generalized stress-energy tensor) to gravity — as directions for future work, with consistency requirements such as energy–momentum conservation unexamined (Kent notes in the predecessor paper that a stress-energy tensor beable need not generally satisfy conservation laws everywhere and cannot be directly used as a source in Einstein's equations).

In the direct sequel, Kent replaced the idealized stress-energy measurements by late-time photodetection (Phys. Rev. A 96, 062121 (2017)), using photon wave mechanics and standard photodetection theory. This raised the infrared problem: realistic charged dynamics produces infinitely many arbitrarily soft photons (bremsstrahlung), so a late-time detector may face infinitely many detections. Kent sketched resolutions — per-hypersurface frequency cutoffs tending to zero as the hypersurface recedes, finite-volume detector arrays, or late-time photon momentum measurements combined with the Aharonov–Bergmann–Lebowitz (ABL) rule, which gives probabilities for intermediate measurements conditioned on both a pre-selected initial state and a post-selected final state; this last option drops the light-cone-restricted position-measurement conditioning of the 2015 proposal in favor of a qualitatively different construction rather than refining it — but explicitly deferred their demonstration. Later analyses (Butterfield 2018; Marsh 2018; Butterfield and Marsh 2019; Adlam 2023) examined the framework's locality verdicts, gave expository presentations, explored the arrows of time it implies, and identified a further gap — when no photons scattered from a macroscopic event escape uninterrupted to infinity (shielding), no quasiclassical history is recovered — without resolving any of the flagged technical strands.

Problem Statement

Determine whether Kent's late-time-measurement beable construction is well-defined and yields a quasiclassical one-world picture beyond the original toy models, characterize the spacetime and cosmological conditions under which it applies, and examine whether the natural generalizations suggested in the source are consistent. Here 'quasiclassical' is fixed by an explicit criterion: at a coarse-graining scale the analysis states, the beable configuration must be concentrated with probability close to 1 on histories lying within a stated tolerance of approximately classical equations of motion, and any claim that the toy-model conclusions survive must report that scale and the size of the stochastic corrections it permits. The single overarching objective — establishing (or refuting) the viability of the proposal outside idealized models — connects four linked open questions flagged by the source and its later literature:

(a) Well-definedness: prove, or disprove by explicit counterexample, the existence of the asymptotic limit $\mathrm{Prob}_{\infty}(C_{S_{1}})=\lim_{S\to\infty}\mathrm{Prob}_{S}(C_{S_{1}})$ for realistic dynamics and coarse grainings — in particular for models with wavepacket spreading, recoil, imperfect scattering and soft-photon (bremsstrahlung) production, where the late-time photodetection outcome space may be infinite — where $S\to\infty$ means that the late-time spacelike hypersurface carrying the fictional measurements recedes indefinitely into the future (in the 2017 photodetection version, that the late-time parameter tends to infinity together with any accompanying auxiliary choices such as detector volume and frequency cutoff). Establish in particular whether the limit is independent of these auxiliary choices, including the choice of the receding hypersurfaces or limit sequence: the predecessor paper defines the beables themselves by a limit taken "assuming both that this limit exists and that it is independent of the chosen limit sequence", while the 2015 paper assumes only existence for $\mathrm{Prob}_{\infty}$, so the status of the independence requirement is itself part of the question. Establish also whether the infrared-problem resolutions sketched by Kent within the framework (receding frequency cutoffs, finite-volume detector arrays) yield beables independent of the auxiliary choices; his third sketch — late-time photon momentum measurements combined with the ABL rule — replaces the light-cone-restricted conditioning of the 2015 proposal by a different construction rather than resolving the infrared problem within it.

(b) Robustness: construct and analyze more realistic 3+1-dimensional models — whose realism must meet the minimum bar of including the features Kent lists: spreading wavepackets, imperfect scattering with reaction on the massive systems, many photons, nonzero Hamiltonians for massive subsystems — and determine whether the qualitative conclusions of the toy models survive, i.e. whether the beables still give a quasiclassical description of reality, including in the cases where scattered photons are shielded or delayed from escaping to future infinity (the gap identified by Butterfield and Marsh; the amendment they outline, which additionally conditions on mass-energy on the future light cone, would modify rather than test the original construction, whose defining rule conditions only on data outside the future light cone). A 3+1-dimensional re-instantiation of the original toy idealizations (perfect instantaneous reflection, $H=0$ for massive subsystems, no bremsstrahlung) does not test strand (b), since it reproduces the idealizations that strand is about.

(c) Domain of validity: characterize the asymptotic conditions on background spacetimes and cosmologies under which massless records escape to future infinity so that the prescription defines useful quasiclassical beables; the source notes that the definitions extend to background spacetimes with appropriate asymptotic properties in the far future and that they fail, for example, in final big-crunch spacetimes, but does not fix a precise criterion for such a characterization.

(d) Generalizations: explore the natural generalizations suggested in the source — modified theories that reweight the beable configuration probabilities, or that couple the stress-energy-type beables to gravity (for example via graviton beables) — and examine their consistency (in particular energy–momentum conservation, given that a stress-energy-tensor beable need not generally satisfy conservation laws everywhere and cannot be directly used as a source in Einstein's equations) together with their empirical consequences; the source treats a full quantum theory of gravity as beyond the proposal's preconditions, noting that without it convincing simple-model illustrations of graviton beables are hard to give and that exploring the possibility further would be interesting.

Each of (a)–(d) is an independently standing part of this objective: resolving any single strand is a substantive partial result, while the full objective is met only when all four are resolved. Within (a), the existence of the limit, its independence from the auxiliary choices (including the choice of the receding hypersurfaces or limit sequence), and the independence of the two sketched infrared resolutions are separately refutable sub-claims, each of which may be reported on its own.

Full rigor in interacting 3+1-dimensional quantum field theory is acknowledged in the source to hinge on the general unsolved problem of constructing rigorous interacting relativistic quantum field theory; results at the level of rigor of standard effective or perturbative treatments, or within well-specified model classes, are legitimate progress on this problem as posed.

References

1. Adrian Kent, "Lorentzian quantum reality: postulates and toy models", Philosophical Transactions of the Royal Society A 373, 20140241 (2015), DOI 10.1098/rsta.2014.0241; arXiv:1411.2957, https://arxiv.org/abs/1411.2957
2. Adrian Kent, "Solution to the Lorentzian quantum reality problem", Physical Review A 90, 012107 (2014), DOI 10.1103/PhysRevA.90.012107; arXiv:1311.0249, https://arxiv.org/abs/1311.0249
3. Adrian Kent, "Quantum reality via late-time photodetection", Physical Review A 96, 062121 (2017), DOI 10.1103/PhysRevA.96.062121; arXiv:1608.04805, https://arxiv.org/abs/1608.04805
4. Jeremy Butterfield, "Peaceful Coexistence: Examining Kent's Relativistic Solution to the Quantum Measurement Problem", in Reality and Measurement in Algebraic Quantum Theory (Springer Proceedings in Mathematics & Statistics, vol. 261, 2018), pp. 277-314, DOI 10.1007/978-981-13-2487-1_11; arXiv:1710.07844, https://arxiv.org/abs/1710.07844
5. Brendan Marsh, "Depictions of Quantum Reality in Kent's Interpretation of Quantum Theory", arXiv:1811.08950 (2018), https://arxiv.org/abs/1811.08950
6. Jeremy Butterfield and Brendan Marsh, "Non-locality and quasiclassical reality in Kent's formulation of relativistic quantum theory", Journal of Physics: Conference Series 1275, 012002 (2019), DOI 10.1088/1742-6596/1275/1/012002; arXiv:1902.01294, https://arxiv.org/abs/1902.01294
7. Emily Adlam, "Disappearing Without a Trace: The Arrows of Time in Kent's Solution to the Lorentzian Quantum Reality Problem", The British Journal for the Philosophy of Science 77, 373-398 (print 2026; online 2023), DOI 10.1086/726085; arXiv:2305.13201, https://arxiv.org/abs/2305.13201
8. Adrian Kent, "Toy Models of Top Down Causation", Entropy 22(11), 1224 (2020), DOI 10.3390/e22111224; arXiv:1909.12739, https://arxiv.org/abs/1909.12739
9. Adrian Kent, "Environmental Collapse Models", arXiv:2206.02711 (2022), https://arxiv.org/abs/2206.02711
10. Adrian Kent, "Fundamental physics, existential risks and human futures", Philosophical Transactions of the Royal Society A 383, 20230376 (2025), DOI 10.1098/rsta.2023.0376; arXiv:2604.26530, https://arxiv.org/abs/2604.26530