An Open System Approach to Gravity

TL;DR

This paper develops a comprehensive open-system framework for gravity, incorporating quantum and classical effects, to better understand cosmological phenomena like inflation and gravitational wave propagation, including dissipative effects.

Contribution

It introduces a general open gravitational dynamics framework based on general relativity and the Schwinger-Keldysh formalism, extending inflationary effective field theory and analyzing gravitational wave birefringence.

Findings

Reproduces the Open Effective Field Theory of Inflation including gravitational interactions

Derives dissipative and conservative corrections to gravitational wave propagation

Finds gravitational birefringence is primarily dissipative at leading order

Abstract

Several major open problems in cosmology, including the nature of inflation, dark matter, and dark energy, share a common structure: they involve spacetime-filling media with unknown microphysics, and can be probed so far only through their gravitational effects. This observation motivates a systematic open-system approach to cosmology, in which gravity evolves in the presence of a generic, unobservable environment. In this work, we develop a general framework for open gravitational dynamics based on general relativity and the Schwinger-Keldysh formalism, carefully addressing the nontrivial constraints imposed by diffeomorphism invariance. At the quantum level, our path integral formulation computes the gravitational density matrix in perturbation theory around a semi-classical spacetime. As illustrative applications, we study inflation and the propagation of gravitational waves in classical regimes where environmental interactions are non-negligible. In the inflationary context, our framework reproduces the known Open Effective Field Theory of Inflation in the decoupling limit and extends it to include gravitational interactions. For gravitational waves, we derive the most general conservative and dissipative corrections to propagation. Remarkably, we find that the leading-order gravitational birefringence is dissipative in nature, whereas conservative birefringence appears only at higher derivative order, opposite to the electromagnetic case. Our results pave the way to modeling dissipative effects in the late universe.