A general theory of linear cosmological perturbations: bimetric theories
(28 pages)

By considering the most general local quadratic gravitational action for a given field content, a perturbative scheme is constructed which respects any gauge symmetries. This gives a parametrization of the linear cosmological evolution in any gravitational theory* with a particular field content, for example Horndeski, Generalized Proca, EinsteinAether, massive bigravity and Eddingtoninspired Born Infeld.
(*without derivative interactions)

This parametrization provides a useful framework within which to discuss the cosmological predictions of a very large class of bimetric theories. This allows much more efficient comparison with data. In principle, this can also be extended to multimetric theories of gravity.

Causality Constraints on Massive Gravity
(4 pages)

It was recently shown that only a finite region of parameter space in dRGT massive gravity is capable of admitting a unitary, analytic UV completion. This paper considers a ppwave background (written in KerrSchild form), and shows that there can be a signal time advance which is within the EFT regime of validity, unless the dRGT parameter α_{3}=1/2. They argue that this implies the existence of closed timelike curves, and therefore that one should further restrict the dRGT parameter space to this line.

The low energy effective description of dRGT massive gravity involves only two free parameters once one imposes the required symmetries and classical ghost freedom. By demanding additional properties of the theory, one can constraint these free parameters, which is useful for future phenomenological investigations (as then only a finite region of parameter space needs to be tested against observation in order to rule the theory out). The additional property demanded here is the absence of closed timelike curves on a particular nontrivial background. However, while the authors succeed in showing that there is a resolvable time advance, this a weaker observation than closed timelike curves. Also, the background itself saturates the Null Energy Condition, is not asymptotically flat, and contains a singularity, and so is not very physical. One may then wonder whether we should take the existence of these time advances as particulary dangerous for the theory.
