Do baby universes and α-parameters exist in a complete theory of quantum gravity? The open status of Hawking's baby-universe framework

Stephen Hawking's lecture 'Baby Universes' (Mod. Phys. Lett. A 5 (1990) 145), which consolidated the late-1980s claim that Planck-scale closed universes branching off from spacetime induce an infinite set of effective interactions on low-energy physics whose strengths, in baby-universe coherent states (q-states), become undetermined coupling constants α — rendering quantum gravity effectively non-renormalizable even if the underlying theory is finite, equating apparent loss of quantum coherence with initial-state uncertainty, and, combined with the no-boundary wave function, predicting a cosmological constant $\Lambda = q m_p^2 + \Lambda_0$ with probability $e^{-q^2} e^{3\pi m_p^2/\Lambda}$ concentrated at $\Lambda = 0$. After a 2020–2026 revival through the black-hole information problem and holography, mutually incompatible results — ensemble/baby-universe field theory (Marolf–Maxfield), swampland one-dimensionality (McNamara–Vafa), stringy disorder averaging (Heckman–Turner–Yu), no-go theorems for semiclassical baby universes in large-N AdS/CFT (Engelhardt–Gesteau, Gesteau), a large-N filter under which internal wormholes do not induce random couplings (Liu), and third-quantized emergence without α-parameters (Harlow) — all remain restricted to controlled lower-dimensional or large-N models, while the no-boundary proposal itself is still contested in 2026. The surviving open problem is to determine, within a complete unitary theory of quantum gravity such as string theory or a holographic dual, whether baby-universe topology change and α-parameters exist, what this implies for the predictability of low-energy couplings and for quantum coherence, and whether the no-boundary state then drives the cosmological constant to zero.

Background

In general relativity the topology of spacetime is fixed classically, but a quantum theory of gravity is expected to sum over topologies in its path integral. In the late 1980s Hawking, Coleman, Giddings, and Strominger argued that this sum includes Euclidean wormholes — narrow 'handles' (computed in the Wick-rotated, imaginary-time form of the theory) that connect distant points or pinch off entirely — whose mouths appear in the large universe as the creation and absorption of Planck-scale closed universes, called baby universes. Because no observer in the large universe can measure the internal state of a baby universe, tracing over these degrees of freedom seems to turn pure quantum states into mixed ones; Giddings and Strominger called this loss of incoherence and showed that it can be repackaged as an indeterminacy of coupling constants. In the equivalent third-quantized picture, one promotes whole universes to creation and annihilation operators $a_i$, $a_i^\dagger$ (one pair per baby-universe species $i$); the effect on low-energy physics is then an infinite set of effective interactions of the form (a local polynomial in the ordinary fields) times $(a_i + a_i^\dagger)$. In a coherent eigenstate of $(a_i + a_i^\dagger)$ — a q-state with eigenvalue $\sqrt{2} q_i$ — these become ordinary local interactions with couplings $\alpha_i = \sqrt{2} q_i$, the α-parameters. Hawking's lecture 'Baby Universes' computed these vertices explicitly, argued that the empty baby universe shifts the cosmological constant to $\Lambda = q m_p^2 + \Lambda_0$ ($m_p$ is the Planck mass, $\Lambda_0$ the bare value), and combined the q-state distribution with the no-boundary proposal of Hartle and Hawking — the boundary condition defining the wave function of the universe by a path integral over compact Euclidean geometries with no other boundary — to claim that the probability for the observed cosmological constant, proportional to $e^{-q^2} e^{3\pi m_p^2/\Lambda}$, is overwhelmingly concentrated at $\Lambda = 0$. The lecture also argued that, because the tower of α-parameters is infinite and undetermined, quantum gravity is effectively non-renormalizable — its low-energy theory requires infinitely many measured couplings — even if the underlying fundamental theory, such as string theory (the leading candidate unification of gravity and quantum mechanics), is finite; and that the apparent loss of quantum coherence in black-hole evaporation and baby-universe emission is equivalent to uncertainty about the initial quantum state.

Three lines of criticism emerged immediately and persist: the Fischler–Susskind 'wormhole catastrophe' argues that the same semiclassical calculation predicts an overdensity of large wormholes, breaking the dilute-gas approximation — the assumption that wormholes are sufficiently sparse to be treated as independent — on which the framework rests; the probability density for $\Lambda$ is not normalizable as $\Lambda \to 0^+$; and the Euclidean gravitational path integral itself is not known to be well-defined, since gravity is non-renormalizable and the sum over topologies diverges. A 2018 review by Hebecker, Mikhail, and Soler concluded that whether topology change at low energies is required or forbidden in four and higher dimensions remains to be conclusively settled, and that the conflict between wormholes in AdS string compactifications and the boundary interpretation of α-parameters remains to be understood.

The question returned to the center of the field from 2020 through the black-hole information problem: entropy calculations for Hawking radiation involve replica wormholes — saddle geometries with multiple boundaries that reproduce the unitary Page curve (the radiation entropy rising and then falling as evaporation completes) — and these same multi-boundary topologies conflict with holography. In holography (AdS/CFT), quantum gravity in an asymptotically anti-de Sitter bulk spacetime is exactly dual to a non-gravitational conformal field theory on the boundary, so a wormhole connecting two boundaries would make the product of two independent boundary partition functions fail to factorize, contradicting the independence of the two boundary theories — the factorization puzzle. McNamara and Vafa argued from swampland consistency conditions (conjectured restrictions on which low-energy effective theories admit an ultraviolet completion into quantum gravity) that the baby-universe Hilbert space must be one-dimensional, forbidding α-parameters, while Marolf and Maxfield developed the opposite, ensemble-based picture in which the boundary theory's Hilbert-space dimension itself becomes random. Subsequent no-go theorems and alternative proposals through 2026, tested mainly in solvable settings — two-dimensional JT (Jackiw–Teitelboim) gravity, a model in which gravity in two spacetime dimensions couples to a single scalar field (the dilaton); topological models, which have only global rather than local degrees of freedom; and large-N limits, in which the number of boundary degrees of freedom labeled by N is taken to infinity so that the bulk becomes increasingly classical — have sharpened but not settled which picture is correct for four-dimensional gravity and for our universe; the status of the no-boundary proposal as a fundamental principle is separately contested.

Problem Statement

Determine whether Hawking's baby-universe framework is realized in a complete, consistent theory of quantum gravity, with the same generality as the original claim. Concretely, answer the following jointly linked questions:

(1) Existence. Does a complete unitary theory of quantum gravity — string theory or a holographic dual — admit baby-universe topology change and the associated α-parameters, i.e. coherent q-states of the closed-universe sector whose eigenvalues $\sqrt{2} q_i$ of $(a_i + a_i^\dagger)$ act as the low-energy coupling constants, with effective interactions of the form (local operator) $\times (a_i + a_i^\dagger)$? Or is the closed-universe sector instead trivial (one-dimensional, with no α-parameters)? Distinctly from a trivial sector, the wormhole contributions to the gravitational path integral might alternatively be explained by ensemble averaging (in which α-type structure of the closed-universe sector survives but a single factorizing theory is abandoned), by a large-N filter, or by observer-dependent rules.

(2) Predictivity and coherence. If α-parameters exist: are the low-energy couplings then genuinely undetermined up to their probability distribution (Hawking's effective non-renormalizability — an infinite set of interactions whose strengths the theory does not fix and that must therefore be measured, with the probability of large deviations Planck-suppressed), and what physical principle or initial state fixes that distribution and selects the observed values of the scalar masses, gauge couplings, and the cosmological constant? If α-parameters do not exist: what explains the topology-changing contributions to multi-boundary gravitational path integrals (the factorization puzzle), and is the dichotomy 'loss of quantum coherence versus initial-state uncertainty' thereby dissolved in favor of exact unitarity?

(3) Cosmological constant. Given the answers to (1) and (2), and taking the no-boundary (Hartle–Hawking) state as the framework's assumed input: does it imply the claimed q-state distribution $e^{-q^2}$ together with the enhancement $e^{3\pi m_p^2/\Lambda}$ that concentrates $\Lambda = q m_p^2 + \Lambda_0$ at zero?

These three questions are one scientific objective — the consistency and empirical content of the baby-universe mechanism inside a complete theory — and must be answered together: (2) is vacuous without (1), (3) presupposes both, and results obtained in controlled models (one- or two-dimensional gravity, JT gravity, topological models, large-N limits) count toward resolution only insofar as they are extended to four-dimensional gravity, including the cosmological setting to which the $\Lambda = 0$ prediction was applied, or are accompanied by a controlled account of why the mechanism is special to the model. The problem is resolved when, for a specified complete theory or class of theories, one can state whether baby universes and α-parameters exist, what this implies for the predictability of low-energy physics, for quantum coherence, and for the cosmological constant, and why the semiclassical wormhole signatures of the gravitational path integral take the form they do.

References

1. S. W. Hawking, 'BABY UNIVERSES', Modern Physics Letters A 5 (1990) 145–155, DOI 10.1142/s0217732390000184, https://doi.org/10.1142/s0217732390000184
2. S. W. Hawking and R. Laflamme, 'Baby universes and the non-renormalizability of gravity', Physics Letters B 209 (1988) 39–41, DOI 10.1016/0370-2693(88)91825-4, https://doi.org/10.1016/0370-2693(88)91825-4
3. Sidney Coleman, 'Why there is nothing rather than something: A theory of the cosmological constant', Nuclear Physics B 310 (1988) 643–668, DOI 10.1016/0550-3213(88)90097-1, https://doi.org/10.1016/0550-3213(88)90097-1
4. Steven B. Giddings and Andrew Strominger, 'Axion-induced topology change in quantum gravity and string theory', Nuclear Physics B 306 (1988) 890–907, DOI 10.1016/0550-3213(88)90446-4, https://doi.org/10.1016/0550-3213(88)90446-4
5. Steven B. Giddings and Andrew Strominger, 'Loss of incoherence and determination of coupling constants in quantum gravity', Nuclear Physics B 307 (1988) 854–866, DOI 10.1016/0550-3213(88)90109-5, https://doi.org/10.1016/0550-3213(88)90109-5
6. Steven B. Giddings and Andrew Strominger, 'Baby universe, third quantization and the cosmological constant', Nuclear Physics B 321 (1989) 481–508, DOI 10.1016/0550-3213(89)90353-2, https://doi.org/10.1016/0550-3213(89)90353-2
7. S. W. Hawking, 'BABY UNIVERSES II', Modern Physics Letters A 5 (1990) 453–466, DOI 10.1142/S0217732390000524, https://doi.org/10.1142/S0217732390000524
8. W. Fischler and L. Susskind, 'A wormhole catastrophe', Physics Letters B 217 (1989) 48–54, DOI 10.1016/0370-2693(89)91514-1, https://doi.org/10.1016/0370-2693(89)91514-1
9. Job Feldbrugge, Jean-Luc Lehners, and Neil Turok, 'No smooth beginning for spacetime', Physical Review Letters 119 (2017) 171301, DOI 10.1103/PhysRevLett.119.171301, https://arxiv.org/abs/1705.00192
10. Arthur Hebecker, Thomas Mikhail, and Pablo Soler, 'Euclidean wormholes, baby universes, and their impact on particle physics and cosmology', Frontiers in Astronomy and Space Sciences 5 (2018) 35, DOI 10.3389/fspas.2018.00035, https://arxiv.org/abs/1807.00824
11. Donald Marolf and Henry Maxfield, 'Transcending the ensemble: baby universes, spacetime wormholes, and the order and disorder of black hole information', JHEP 08 (2020) 044, DOI 10.1007/JHEP08(2020)044, https://arxiv.org/abs/2002.08950
12. Jacob McNamara and Cumrun Vafa, 'Baby Universes, Holography, and the Swampland', arXiv:2004.06738, https://arxiv.org/abs/2004.06738
13. Jonathan J. Heckman, Andrew P. Turner, and Xingyang Yu, 'Disorder averaging and its UV discontents', Physical Review D 105 (2022) 086021, DOI 10.1103/PhysRevD.105.086021, https://arxiv.org/abs/2111.06404
14. Norihiro Iizuka, Akihiro Miyata, and Tomonori Ugajin, 'A comment on a fine-grained description of evaporating black holes with baby universes', JHEP 09 (2022) 158, DOI 10.1007/JHEP09(2022)158, https://arxiv.org/abs/2111.07107
15. Jean-Luc Lehners, 'Review of the no-boundary wave function', Physics Reports 1022 (2023) 1–82, DOI 10.1016/j.physrep.2023.06.002, https://arxiv.org/abs/2303.08802
16. Netta Engelhardt and Elliott Gesteau, 'Further Evidence Against a Semiclassical Baby Universe in AdS/CFT', arXiv:2504.14586, https://arxiv.org/abs/2504.14586
17. Elliott Gesteau, 'A no-go theorem for large N closed universes', arXiv:2509.14338, https://arxiv.org/abs/2509.14338
18. Hong Liu, 'Filtering CFTs at large N: Euclidean Wormholes, Closed Universes, and Black Hole Interiors', arXiv:2512.13807, https://arxiv.org/abs/2512.13807
19. Daniel Harlow, 'Observers, α-parameters, and the Hartle-Hawking state', arXiv:2602.03835, https://arxiv.org/abs/2602.03835
20. Ahmed I. Abdalla, Stefano Antonini, Raphael Bousso, Luca V. Iliesiu, Adam Levine, and Arvin Shahbazi-Moghaddam, 'Consistent Evaluation of the No-Boundary Proposal', arXiv:2602.02682, https://arxiv.org/abs/2602.02682
21. Ying Zhao, 'It from Bit: The Hartle-Hawking state and quantum mechanics for de Sitter observers', arXiv:2602.05939, https://arxiv.org/abs/2602.05939
22. Suzanne Bintanja and Elliott Gesteau, 'The α-states of a string worldsheet', arXiv:2607.15336, https://arxiv.org/abs/2607.15336
23. Shinji Hirano, 'A Third-Quantized Description of Spacetime Wormholes in AdS/CFT', arXiv:2608.23111, https://arxiv.org/abs/2608.23111