Science › Creation theories
The no-boundary proposal (Hartle–Hawking)
1983· Theory
SpeculativeA serious idea with no experimental proof yet.
James Hartle and Stephen Hawking's 1983 proposal that the universe's past is finite but has no edge, so it needs no starting conditions.
See it on the Science map →How it works
Quantum physics computes odds by adding up every possible history. Hartle and Hawking applied that to the whole universe, but counted only histories that close off smoothly, with no boundary. Near the start, time turns into a fourth direction of space (physicists call it 'imaginary time'), so the earliest universe is a smooth four-dimensional cap, shaped a bit like the rounded end of a badminton shuttlecock. Ordinary flowing time appears as that cap widens into an expanding universe.
Why it matters
It offered a universe that explains its own beginning, with no singularity and nothing set from outside. It became the best-known quantum picture of creation.
How we know
None. It is a mathematical rule, and calculations disagree on what it predicts: early versions seemed to favor universes with little inflation, at odds with the one we observe.
Still a mystery
Whether its math is well defined, and whether it needs a multiverse or observers to give sensible predictions, are still argued. Hawking was refining it with Thomas Hertog until his death in 2018.
Connections (4)
▸Stephen Hawking proposed with James Hartle The no-boundary proposal (Hartle–Hawking)Speculative
The relationship
Hawking brought the cosmology; Hartle, at UC Santa Barbara, brought quantum methods for whole systems. Together they wrote down a specific 'wave function of the universe', the first concrete rule for its quantum starting state. Hawking then popularized it, writing in 1988 that a universe with no boundary would simply be, and asking what place that left for a creator.
How the link was found
Hawking first floated it at a 1981 Vatican conference on cosmology, where, he recalled, Pope John Paul II urged scientists not to probe the beginning itself. Hartle and Hawking published 'Wave function of the Universe' in 1983.
Is it proven?
Untested. From the 2000s Hawking developed it with Thomas Hertog into a 'top-down' view, in which the present universe selects its likely histories. James Hartle (1939–2023) continued the work too.
Open questions and mysteries
Did Hawking's final, observer-centered version improve the proposal or abandon its original simplicity? Colleagues still disagree.
▸The no-boundary proposal (Hartle–Hawking) says "there is no before" to What came before the Big Bang?Speculative
The relationship
The past is finite but has no first instant and no singularity. Shrink the universe backward and, near the start, time smoothly becomes space, so the history closes off instead of ending at an edge. With no edge, nothing has to be set from outside; the laws alone fix the starting state. 'Before' simply has nothing to refer to.
How the link was found
Hartle and Hawking, 1983, a year after Vilenkin's rival tunneling proposal. Both aimed to make the beginning itself part of physics rather than an unexplained input.
Is it proven?
No observation yet tests it. Its predictions for inflation and the microwave-background ripples depend on contested calculations, so it can be neither confirmed nor ruled out.
Open questions and mysteries
Does 'no before' answer the question or dissolve it? And does it produce a smooth start like ours, or a different universe?
▸The no-boundary proposal (Hartle–Hawking) makes time emerge, reframing What is time, really?Speculative
The relationship
In the no-boundary picture, time is not fundamental. At the very start all four dimensions behave like space; ordinary time, with a past and a future, appears only as the universe expands. That fits the Wheeler–DeWitt equation, which has no time in it: time becomes something that emerges inside the universe, tracked by its own changing size.
How the link was found
Hartle and Hawking (1983). Hawking later used it on the arrow of time, first claiming disorder would decrease if the universe recollapsed; Don Page and Raymond Laflamme convinced him otherwise, and he called it a mistake.
Is it proven?
Unproven. It is mathematically consistent, but no measurement reaches the epoch where time would emerge.
Open questions and mysteries
If time emerges, what does 'the universe began' mean? Why does emergent time have a direction, from the smooth early universe toward disorder?
▸The wave function is applied to the whole universe in The no-boundary proposal (Hartle–Hawking)Speculative
The relationship
Normally a wave function gives odds for a particle's position. The 'wave function of the universe' gives odds for every possible shape and size of space and its fields. The Wheeler–DeWitt equation says how it must behave, but allows many solutions. The no-boundary proposal picks one: add up only histories that round off smoothly with no edge.
How the link was found
Bryce DeWitt wrote the universe's wave equation in 1967. Hartle and Hawking chose its solution in 1983 using Feynman's sum over histories; Vilenkin proposed a rival choice.
Is it proven?
Untested. The calculations work only in simplified models with a few variables. Whether a wave function for everything makes sense is unclear, since no outside observer exists to measure it.
Open questions and mysteries
With no outside observer, what makes one outcome real? This drove Hartle and Murray Gell-Mann's 'decoherent histories' work from 1990, and feeds interest in many-worlds.