Cosmology from the vacuum

TL;DR

The paper develops a holographic framework for cosmology with a negative fundamental cosmological constant $\Lambda<0$, where accelerated expansion can arise from time-dependent scalar fields evolving toward negative potential regions. It shows that cosmology can be described by vacuum physics in a static planar AdS wormhole, with cosmological observables obtainable via analytic continuation from the wormhole vacuum and computed reliably within an effective field theory. The construction relies on a 3D holographic CFT coupled to a smaller 4D CFT (often realized in a 3D–4D–3D setup) to produce a dual geometry and a special time-symmetric state defined by Euclidean path integrals. Fluctuations and perturbations are addressed by relating cosmological correlators to vacuum correlators in the wormhole, offering new routes to explain correlations across causally disconnected regions and discuss cosmological islands, while highlighting challenges in realizing fully microscopic, quantitatively realistic models.

Abstract

We argue that standard tools of holography can be used to describe fully non-perturbative microscopic models of cosmology in which a period of accelerated expansion may result from the positive potential energy of time-dependent scalar fields evolving towards a region with negative potential. In these models, the fundamental cosmological constant is negative, and the universe eventually recollapses in a time-reversal symmetric way. The microscopic description naturally selects a special state for the cosmology. In this framework, physics in the cosmological spacetime is dual to the vacuum physics in a static planar asymptotically AdS Lorentzian wormhole spacetime, in the sense that the background spacetimes and observables are related by analytic continuation. The dual spacetime is weakly curved everywhere, so any cosmological observables can be computed in the dual picture via effective field theory without detailed knowledge of the UV completion or the physics near the big bang. In particular, while inflation may explain the origin of perturbations in the cosmology picture, the perturbations can be deduced from the dual picture without any knowledge of the inflationary potential.

Figures (8)

• Figure 1: Asymptotically AdS regions in the analytic continuation of a time-symmetric $\Lambda < 0$ cosmology. The asymptotically AdS Euclidean gravity theory can be given a microscopic holographic description, and this can be used to define a special state for the cosmology.
• Figure 2: Left: two non-interacting 3D CFTs are dual to a pair of disconnected AdS spacetimes. Right: two 3D CFTs interacting via a non-holographic 4D CFT, proposed to be dual (in some cases) to a Lorentzian wormhole with two asymptotically AdS boundaries.
• Figure 3: Two different slicings of the same microscopic Euclidean path integral give states dual to the Lorentzian wormhole and the cosmology. The pictures on the left and right indicated a Euclidean path integral over field configurations with the specified boundary conditions at the top surface. See also Figure 6 for the gravity interpretation of these states.
• Figure 4: Observables restricted to an $\mathbb{R}^2$ at the time-reflection symmetric surface in the cosmology (black line) are equal to corresponding observables on the $\mathbb{R}^2$ in the middle of the Lorentzian wormhole. More general observables are related by double analytic continuation.
• Figure 5: Effective field theory description of holographic 3D-4D coupled theories with $c_{3D}\gg c_{4D}$. a) Microscopic picture b) Higher-dimensional gravity picture: 4D gravitational physics of the 3D theory appears as the physics of an ETW brane near the AdS boundary. c) Isolating the 5D bulk gravitational physics. d) 5D gravitational physics is equivalent to physics of the dual 4D CFT with a cutoff (indicated by *) on half the space. e) Final effective description: a 4D CFT coupled to a 4D gravity theory + cutoff CFT. The state of the lower 4D CFT in the final picture is different than the state in the micoscopic picture.