Hartle-Hawking No-Boundary

1983James Hartle, Stephen HawkingSpeculativeAlso answers, Ch.02 The Origin of Our Universe5 primary sources, 2 established Reviewed May 14, 2026

The universe has no temporal boundary, time itself emerged from a smooth Euclidean cap.

Visualisation · Quantum Tunneling Origin

§1 · The claim, in one sentence

The universe has no temporal boundary at the beginning. In Euclidean time, the universe is a smooth four-dimensional surface with no edge, like asking what is south of the South Pole.

§2 · Why it might be true

Hartle and Hawking proposed in 1983 that the universe's wavefunction is given by a path integral over all compact Euclidean four-geometries with no boundary. In Euclidean (imaginary) time, the universe is a smooth four-dimensional surface that smoothly caps off, like the rounded bottom of a sphere.

There is no "moment of creation" because there is no temporal boundary. Asking what happened before the Big Bang is, in this framework, like asking what is south of the South Pole. Time emerges as we move away from the Euclidean cap into Lorentzian (real-time) signature.

The family stance

Either literally nothing (Vilenkin), or a state without a temporal boundary (Hartle-Hawking), or a zero-energy quantum fluctuation (Tryon). All deny that a classical "before" makes sense.

§2.5 · Evidence

Elegant mathematical formulation
Eliminates singularity problem
Predicts de Sitter-like beginning compatible with inflation

§3 · What you'd need to test it

No initial singularity
Universe wavefunction smooth at a = 0
Predicts a specific spectrum of cosmological perturbations

§4 · Where it breaks

Feldbrugge, Lehners, Turok (2017) argue the proposal predicts an unstable universe
Direction of "outgoing" vs "incoming" mode is contested
A more careful mathematical analysis of the path integral suggests deeper problems, hinting the proposal may predict unstable outcomes (see Go Deeper)
Maldacena (2024) re-examines the no-boundary proposal and finds it predicts spatial curvature in conflict with observations, and is non-normalizable for landscape-like potentials.
Ivo, Li & Maldacena (2024) show that once you account for the regions of space we cannot observe, the mathematics of the no-boundary state leads to physically unrealistic predictions, sharpening the case against the original proposal.

Go deeper

In the no-boundary proposal, the wavefunction of the universe is given by a Euclidean path integral over compact four-manifolds with the present three-geometry as the only boundary. There is no initial three-geometry; the path integral closes off smoothly at small a.

In Lorentzian signature, the universe emerges from the Euclidean region and evolves classically. The proposal is closely related to Vilenkin's but with a different choice of contour: Hartle-Hawking selects a no-boundary condition, Vilenkin selects an outgoing tunneling condition.

Recent work (Feldbrugge, Lehners, Turok 2017, 2018) using Picard-Lefschetz theory suggests that with proper definition of the path integral, the no-boundary proposal predicts perturbation amplitudes that grow exponentially, an unstable universe. This is hotly debated.

▸§5 · Who built it, and when(5 sources, 2 established, 3 debated)

Variants in this family

Compare variants

0votes

Currently #2 in this family · #6 in Ch.1

Up next

Spotted an error? Have a source to add?

Prefer email?

You can also send a prefilled email with the variant URL already filled in.