Cutoff for extensions of massive gravity and bi-gravity

Andrew Matas

Research output: Contribution to journalArticlepeer-review

13 Scopus citations

Abstract

Recently there has been interest in extending ghost-free massive gravity, bi-gravity, and multi-gravity by including non-standard kinetic terms and matter couplings. We first review recent proposals for this class of extensions, emphasizing how modifications of the kinetic and potential structure of the graviton and modifications of the coupling to matter are related. We then generalize existing no-go arguments in the metric language to the vielbein language in second-order form. We give an ADM argument to show that the most promising extensions to the kinetic term and matter coupling contain a Boulware-Deser ghost. However, as recently emphasized, we may still be able to view these extensions as effective field theories below some cutoff scale. To address this possibility, we show that there is a decoupling limit where a ghost appears for a wide class of matter couplings and kinetic terms. In particular, we show that there is a decoupling limit where the linear effective vielbein matter coupling contains a ghost. Using the insight we gain from this decoupling limit analysis, we place an upper bound on the cutoff for the linear effective vielbein coupling. This result can be generalized to new kinetic interactions in the vielbein language in second-order form. Combined with recent results, this provides a strong uniqueness argument on the form of ghost-free massive gravity, bi-gravity, and multi-gravity.

Original languageEnglish (US)
Article number075004
JournalClassical and Quantum Gravity
Volume33
Issue number7
DOIs
StatePublished - Mar 1 2016
Externally publishedYes

Bibliographical note

Publisher Copyright:
© 2016 IOP Publishing Ltd.

Keywords

  • effective field theory
  • massive gravity
  • modified gravity

Fingerprint

Dive into the research topics of 'Cutoff for extensions of massive gravity and bi-gravity'. Together they form a unique fingerprint.

Cite this