Bidirectional arrows of time in the bouncing universes of the no-boundary quantum state: establishing the result beyond the semiclassical minisuperspace derivation

Hartle and Hertog (Phys. Rev. D 85, 103524 (2012)) showed, using saddle-point (fuzzy-instanton) methods in a minisuperspace model, that the no-boundary wave function (NBWF) of the universe predicts bouncing cosmological histories in which fluctuations are small at the bounce and grow away from it on both sides, so that the fluctuation, thermodynamic, electromagnetic and psychological arrows of time are bidirectional across a bounce - an account of time's arrow from the quantum state alone, without a prescribed low-entropy boundary condition. The derivation rests on four explicitly acknowledged assumptions: classical histories are extrapolated through the region near the bounce where the WKB classicality condition fails; histories are not continued through classically singular regimes; predictions are not observationally controlled in the presence of the large superhorizon inhomogeneities produced by eternal-inflation regimes near the bounce; and nonlinear backreaction beyond linear perturbations about homogeneous isotropic backgrounds is neglected. The problem is to establish or refute the bidirectional-arrow prediction in the quantum theory proper - lifting the semiclassical, truncation, and continuation assumptions. A related observational question - the time-distance of our surface from the small-fluctuation regime at the bounce - is flagged by the authors as a limitation that, in their words, does not alter the arrow-reversal finding.

Background

The no-boundary wave function (NBWF) of Hartle and Hawking (1983) is a proposal for the quantum state of the universe: the wave function of a closed three-geometry is given by a gravitational path integral over compact four-geometries having no boundary other than the final three-surface. In regions of superspace (the configuration space of three-geometries and fields) where the action varies rapidly, the integral can be evaluated by saddle points. The relevant saddles are 'fuzzy instantons' - complex, generally non-real solutions of the classical equations that are regular on the compact disk defining the state - and they give the wave function a WKB (semiclassical) form $\Psi \approx \exp(-I_R + iS)$, where $I_R$ and $S$ are the real and imaginary parts of the Euclidean action. Where the classicality condition holds - the phase $S$ varies rapidly as a function of the boundary data compared to the amplitude $I_R$ - the state predicts an ensemble of classical Lorentzian spacetime histories (the integral curves of $S$) with probabilities proportional to $\exp(-2I_R)$. This is the route by which classical spacetime emerges from quantum gravity in the no-boundary program.

Hartle and Hertog (2012) applied this framework in a minisuperspace model - the scale factor $b$ and homogeneous scalar-field value $\chi$ of a closed, spatially homogeneous, isotropic universe, plus linear perturbation modes $z_{(n)}$ - to derive the arrows of time predicted by the NBWF. Regularity of the perturbed fuzzy instantons at the 'South Pole' (the center of the compact disk, where the universe has zero size) forces fluctuations to vanish there. This implies that each classical Lorentzian history has exactly one range of three-surfaces on which fluctuations are small: at the bounce for histories that reach a minimum radius and re-expand, or near one singularity for recollapsing histories with big bang and big crunch. Fluctuations then grow away from that range - freezing at horizon exit during inflation and growing during matter domination - defining a fluctuation arrow of time. Because free electromagnetic radiation is itself a fluctuation, and because gravitational clumping raises entropy, the electromagnetic and thermodynamic arrows align with the fluctuation arrow, and the psychological arrow (the asymmetry of information gathering and processing) follows from these. In bouncing histories - which, for large values of the South-Pole field $\phi_0$, inflate on both sides of the bounce - the arrows of time are therefore bidirectional: they point away from the bounce on both sides, and events on the far side are causally disconnected from ours.

This gives an explanation of time's arrow from the quantum state alone, without imposing a global low-entropy 'past hypothesis' boundary condition. It contrasts with bidirectional-arrow scenarios constructed outside quantum cosmology (Aguirre and Gratton's null-boundary proposal; Carroll and Chen's spontaneous-inflation model), with earlier discussions of arrows in recollapsing quantum universes (Kiefer and Zeh), and with the analysis of Bojowald in loop quantum cosmology - the application of loop quantum gravity's quantization techniques to homogeneous cosmological models - who reached similar conclusions for quantum bounces in a different quantization.

The 2012 derivation rests on four assumptions the authors explicitly flag. (i) Near a bounce the classicality condition fails (there $S$ varies slowly compared to $I_R$), and the histories are carried through that region by classical extrapolation using the classical equations of motion - 'an assumption which can in principle be checked in the full quantum mechanical theory'. (ii) Histories whose classical evolution is singular (energy density above the Planck density) are not continued; whether history can be extended quantum mechanically through classical singularities is called 'still an open question'. (iii) In regimes of eternal inflation - inflation that perpetually reproduces itself through quantum fluctuations of the field - near the bounce (where $\mathcal{V}^3 > \mathcal{V}_{,\phi}^2$ for the potential $\mathcal{V}$), large superhorizon inhomogeneities (fluctuations on scales larger than the cosmological horizon) form on both sides, making it difficult to estimate how far in time our present surface is from the regime of small fluctuations at the bounce. (iv) The model is a truncation to homogeneous isotropic backgrounds plus linear perturbations with small backreaction; the continuation of the arrow into the nonlinear regime is stated as an expectation, not a result.

Later work substantially developed the foundations of this program: a Lorentzian, Picard-Lefschetz (steepest-descent in the complex lapse) formulation of the NBWF; an exactly evaluated no-boundary path integral in biaxial Bianchi IX minisuperspace (Bianchi IX being the homogeneous but anisotropic closed cosmology in which the single isotropic scale factor is replaced by three independent ones, 'biaxial' meaning two of them are set equal) including the nonlinear backreaction of perturbations; quantum-transition calculations connecting classical patches across bounces and singularities; and a unitary Schrödinger evolution of the minisuperspace NBWF in which the proper time of an observer is introduced as a test-field degree of freedom (Piazza and Vareilles 2026). At the same time, a 2017-2019 controversy over the convergence and sign problem of the Euclidean no-boundary path integral left the semiclassical foundation itself only partially settled, as documented in Lehners' 2023 review, which also cautions that all controlled results to date live in minisuperspace - restricted degrees of freedom that could in principle be spoiled by the modes that were omitted. What remains is precisely the question the 2012 authors raised: does the bidirectional-arrow prediction survive when its assumptions are lifted?

Problem Statement

Determine, within the quantum theory of the no-boundary state rather than its semiclassical saddle-point approximation, whether the arrows of time in its bouncing histories are bidirectional - fluctuations in their ground state near the bounce and growing away from it on both sides, with the electromagnetic, thermodynamic, and psychological arrows aligned with that growth - by lifting the acknowledged assumptions of the 2012 derivation on which that conclusion rests. The common objective uniting items (a), (b) and (d) is to put that derivation's conclusion on quantum-mechanical ground; these items are the assumptions of one calculation, not independent goals. Item (c) records a related observational limitation flagged by the authors - one which the 2012 paper states does not alter the arrow-reversal finding - and is not a required component of an answer to the core question:

(a) Classicality through the bounce. In the region near the bounce where the classicality condition fails (the phase $S$ of the Euclidean action varying slowly compared to its real part $I_R$), determine by quantum evolution of the state - not by classical extrapolation - which classical histories emerge on the far side, whether they form the same ensemble as the classically extrapolated ones, and whether the far-side fluctuation state is the continuation of the near-bounce ground state (so that the fluctuation arrow indeed points away from the bounce on both sides).

(b) Singular regimes. For classical histories that are singular rather than bouncing, determine whether and how the no-boundary state continues histories quantum mechanically through the classically singular region, and whether the 'one range of small fluctuations per history' structure - hence a single arrow direction on each side - survives that continuation.

(c) Observational control. In the presence of the large superhorizon inhomogeneities generated in eternal-inflation regimes near the bounce, quantify what present or future observations can determine about the time-distance of our surface from the small-fluctuation regime at the bounce - measures, bounds, or probabilities for observables, or a demonstration that no observable constrains it.

(d) Backreaction and truncation. Determine the effect of nonlinear backreaction of the fluctuations and of degrees of freedom beyond the homogeneous-isotropic-plus-linear-perturbation truncation on the fluctuation arrow and on the fate of the bounce.

References

1. James Hartle and Thomas Hertog, 'Arrows of time in the bouncing universes of the no-boundary quantum state', Physical Review D 85, 103524 (2012). doi:10.1103/PhysRevD.85.103524; arXiv:1104.1733. https://doi.org/10.1103/PhysRevD.85.103524
2. James B. Hartle and S. W. Hawking, 'Wave function of the Universe', Physical Review D 28, 2960-2975 (1983). doi:10.1103/PhysRevD.28.2960. https://doi.org/10.1103/PhysRevD.28.2960
3. James Hartle and Thomas Hertog, 'Quantum transitions between classical histories: Bouncing cosmologies', Physical Review D 92, 063509 (2015). doi:10.1103/PhysRevD.92.063509; arXiv:1502.06770. https://doi.org/10.1103/PhysRevD.92.063509
4. Sebastian F. Bramberger, Thomas Hertog, Jean-Luc Lehners and Yannick Vreys, 'Quantum transitions through cosmological singularities', Journal of Cosmology and Astroparticle Physics 2017 (07), 007 (2017). doi:10.1088/1475-7516/2017/07/007; arXiv:1701.05399. https://doi.org/10.1088/1475-7516/2017/07/007
5. J. Diaz Dorronsoro, J. J. Halliwell, J. B. Hartle, T. Hertog and O. Janssen, 'Real no-boundary wave function in Lorentzian quantum cosmology', Physical Review D 96, 043505 (2017). doi:10.1103/PhysRevD.96.043505; arXiv:1705.05340. https://doi.org/10.1103/PhysRevD.96.043505
6. J. Diaz Dorronsoro, J. J. Halliwell, J. B. Hartle, T. Hertog, O. Janssen and Y. Vreys, 'Damped perturbations in the no-boundary state', Physical Review Letters 121, 081302 (2018). doi:10.1103/PhysRevLett.121.081302; arXiv:1804.01102. https://doi.org/10.1103/PhysRevLett.121.081302
7. Oliver Janssen, Jonathan J. Halliwell and Thomas Hertog, 'The no-boundary proposal in biaxial Bianchi IX minisuperspace', Physical Review D 99, 123531 (2019). doi:10.1103/PhysRevD.99.123531; arXiv:1904.11602. https://doi.org/10.1103/PhysRevD.99.123531
8. Jonathan J. Halliwell, James B. Hartle and Thomas Hertog, 'What is the no-boundary wave function of the Universe?', Physical Review D 99, 043526 (2019). doi:10.1103/PhysRevD.99.043526; arXiv:1812.01760. https://doi.org/10.1103/PhysRevD.99.043526
9. Job Feldbrugge, Jean-Luc Lehners and Neil Turok, 'No smooth beginning for spacetime', Physical Review Letters 119, 171301 (2017). doi:10.1103/PhysRevLett.119.171301; arXiv:1705.00192. https://doi.org/10.1103/PhysRevLett.119.171301
10. Alice Di Tucci, Job Feldbrugge, Jean-Luc Lehners and Neil Turok, 'Quantum incompleteness of inflation', Physical Review D 100, 063517 (2019). doi:10.1103/PhysRevD.100.063517; arXiv:1906.09007. https://doi.org/10.1103/PhysRevD.100.063517
11. Jean-Luc Lehners, 'Review of the no-boundary wave function', Physics Reports 1022, 1-82 (2023). doi:10.1016/j.physrep.2023.06.002; arXiv:2303.08802. https://doi.org/10.1016/j.physrep.2023.06.002
12. Alexander Vilenkin, 'Arrows of time and the beginning of the universe', Physical Review D 88, 043516 (2013). doi:10.1103/PhysRevD.88.043516; arXiv:1305.3836. https://doi.org/10.1103/PhysRevD.88.043516
13. Federico Piazza and Simeon Vareilles, 'Time as a test-field: the no-boundary Universe in motion and a smooth radiation bounce', Classical and Quantum Gravity 43, 055007 (2026). doi:10.1088/1361-6382/ae49de; arXiv:2505.08703. https://doi.org/10.1088/1361-6382/ae49de
14. Marco de Cesare, 'Arrows of time in bouncing cosmologies', chapter in Time and Timelessness in Fundamental Physics and Cosmology, eds. Silvia De Bianchi, Marco Forgione and Laura Marongui, Fundamental Theories of Physics (Springer, 2024), pp. 253-267. doi:10.1007/978-3-031-61860-4_12; arXiv:2405.01380. https://doi.org/10.1007/978-3-031-61860-4_12
15. Sean M. Carroll, Nadiia Diachenko and Saakshi Dulani, 'Toward a phenomenologically acceptable quantum cyclic universe', arXiv:2605.30405 (2026). https://arxiv.org/abs/2605.30405
16. James B. Hartle, 'The quantum mechanical arrows of time', in Quantum Theory: A Two-Time Success Story (Springer, 2014), pp. 113-128. doi:10.1007/978-88-470-5217-8_8; arXiv:1301.2844. https://doi.org/10.1007/978-88-470-5217-8_8
17. James B. Hartle, 'Arrows of time and initial and final conditions in the quantum mechanics of closed systems like the universe', arXiv:2002.07093 (2020). https://arxiv.org/abs/2002.07093
18. M. Bojowald, 'A momentous arrow of time', in The Arrows of Time, eds. Laura Mersini-Houghton and Rudy Vaas (Springer, 2012), pp. 169-189. doi:10.1007/978-3-642-23259-6_9; arXiv:0910.3200. https://doi.org/10.1007/978-3-642-23259-6_9
19. C. Kiefer and H. D. Zeh, 'Arrow of time in a recollapsing quantum universe', Physical Review D 51, 4145-4153 (1995). doi:10.1103/PhysRevD.51.4145. https://doi.org/10.1103/PhysRevD.51.4145
20. Anthony Aguirre and Steven Gratton, 'Inflation without a beginning: a null boundary proposal', Physical Review D 67, 083515 (2003). doi:10.1103/PhysRevD.67.083515. https://doi.org/10.1103/PhysRevD.67.083515
21. Sean M. Carroll and Jennifer Chen, 'Spontaneous inflation and the origin of the arrow of time', arXiv:hep-th/0410270 (2004). https://arxiv.org/abs/hep-th/0410270
22. James Hartle, S. W. Hawking and Thomas Hertog, 'No-boundary measure in the regime of eternal inflation', Physical Review D 82, 063510 (2010). doi:10.1103/PhysRevD.82.063510; arXiv:1001.0262. https://doi.org/10.1103/PhysRevD.82.063510
23. James Hartle, S. W. Hawking and Thomas Hertog, 'Local observation in eternal inflation', Physical Review Letters 106, 141302 (2011). doi:10.1103/PhysRevLett.106.141302; arXiv:1009.2525. https://doi.org/10.1103/PhysRevLett.106.141302