Towards the most general scalar-tensor theories of gravity: a unified approach in the language of differential forms

TL;DR

This paper systematically classifies scalar-tensor theories of gravity using differential forms, identifying a minimal basis, discovering a new Lagrangian, and clarifying relations among second order theories.

Contribution

It introduces a unified differential forms framework for scalar-tensor theories, revealing a new second order Lagrangian and clarifying the structure of known theories.

Findings

Identified a minimal basis for scalar-tensor theories in any dimension.

Discovered a new second order Lagrangian, kinetic Gauss-Bonnet.

Clarified relations between different second order scalar-tensor theories.

Abstract

We use a description based on differential forms to systematically explore the space of scalar-tensor theories of gravity. Within this formalism, we propose a basis for the scalar sector at the lowest order in derivatives of the field and in any number of dimensions. This minimal basis is used to construct a finite and closed set of Lagrangians describing general scalar-tensor theories invariant under Local Lorentz Transformations in a pseudo-Riemannian manifold, which contains ten physically distinct elements in four spacetime dimensions. Subsequently, we compute their corresponding equations of motion and find which combinations are at most second order in derivatives in four as well as arbitrary number of dimensions. By studying the possible exact forms (total derivatives) and algebraic relations between the basis components, we discover that there are only four Lagrangian combinations producing second order equations, which can be associated with Horndeski's theory. In this process, we identify a new second order Lagrangian, named kinetic Gauss-Bonnet, that was not previously considered in the literature. However, we show that its dynamics is already contained in Horndeski's theory. Finally, we provide a full classification of the relations between different second order theories. This allows us to clarify, for instance, the connection between different covariantizations of Galileons theory. In conclusion, our formulation affords great computational simplicity with a systematic structure. As a first step we focus on theories with second order equations of motion. However, this new formalism aims to facilitate advances towards unveiling the most general scalar-tensor theories.