Non-Gaussianity and Strong-Coupling Problem in a Two-Field DHOST Bouncing Model

Ok Song An, Kon Hong, Jin U Kang, and Ui Ri Mun · 2026

Read Non-Gaussianity and Strong-Coupling Problem in a Two-Field DHOST Bouncing Model by Ok Song An, Kon Hong, Jin U Kang, and Ui Ri Mun (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-03302.txt.

Introduction

An, Hong, Kang, and Mun extend a bouncing cosmology model to the nonlinear level. Their two-field DHOST model was already viable at the linear level. It has no BKL instability, no ghost or gradient instability, and no faster-than-light modes. Its scalar spectral index and tensor-to-scalar ratio match observations. The open risk was nonlinear behavior.

First, the team refines the model so its local non-Gaussianity value fits the data. The linear-level viability stays intact through the change. Next, they show the strong-coupling scale stays well above the background energy scale at all times. The model is thus weakly coupled where it must be. It stays stable and non-superluminal while it matches the data. The authors present it as a fully viable two-field DHOST bounce. This one, they argue, does.

Catalog record

CREATOROk Song An, Kon Hong, Jin U Kang, and Ui Ri Mun
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE78 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2606.03302

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Non-Gaussianity and Strong-Coupling Problem in a Two-Field DHOST Bouncing Model Ok Song An, Kon Hong, Jin U Kang, and Ui Ri Mun Abstract We recently constructed a two-field Degenerate Higher-Order Scalar-Tensor (DHOST) bouncing model which is fully viable at the linear level 7 . This model is completely free of Belinski-Khalatnikov-Lifshitz (BKL) instability, ghost instability, gradient instability and superluminality. It also predicts the scalar spectral index and tensor-to-scalar ratio consistent with observations. The aim of this paper is to extend the viability of the model to the non-linear level. To this end, we first refine the original model such that its prediction on the (local) non-Gaussianity parameter f N ​ L f_{NL} agrees with observations, leaving the viability of the model at the linear level intact. We furthermore demonstrate that the strong-coupling scale is well above the characteristic background energy scale all the time. Our model indeed exemplifies the fully viable two-field DHOST bouncing model, in the sense that it is weakly-coupled, stable and non-superluminal as well as consistent with observations. 1 Introduction Bouncing cosmologies (see 60 ; 9 ; 16 for a review) are regarded as a compelling alternative to the inflationary paradigm 27 ; 65 ; 55 , since they provide a natural solution to the initial singularity problem 61 ; 28 that the standard Big Bang cosmology and even the inflationary scenarios 14 ; 13 suffer from (see however 51 ). But within the framework of General Relativity cosmic bounce requires violation of the Null Energy Condition (NEC), which generally leads to various pathologies such as ghost instability, gradient instability and superluminality 64 ; 59 . Modified gravity theories have been widely employed to obtain a non-pathological bouncing scenario. Among these, an exceptional subclass 58 ; 59 of Degenerate Higher-Order Scalar-Tensor (DHOST) Ia theories 44 ; 12 ; 43 ; 47 ; 38 deserves a special attention, since any DHOST theory which can realise an observationally compatible bouncing scenario without the aforementioned pathologies falls into this subclass. However, it was shown in 7 that the exceptional subclass of DHOST Ia theories serves only as a necessary condition, not a sufficient condition for the non-pathological bounce. Upon deriving the sufficient condition, the authors of 7 constructed an explicit bouncing model for the first time which admits a cosmological bouncing solution theoretically healthy and phenomenologically viable in the linear regime , namely • the model is free from not only ghost instability, gradient instability and superluminality but also Belinsky-Khalatnikov-Lifshitz (BKL) instability 11 , • its predictions on the scalar spectral index and tensor-to-scalar ratio are compatible with observations 6 . As we will briefly review in Section 2 , this model comprises of two scalar fields, one of which is responsible for bouncing dynamics and described by the exceptional subclass of DHOST Ia theories with conventional asymptotics in the far past and far future. This scalar field produces blue-tilted scalar perturbations during ekpyrotic contracting phase 36 ; 17 ; 52 ; 20 ; 29 ; 34 ; 31 needed to avoid the BKL instability. It calls for the other scalar field which can generate nearly scale-invariant scalar perturbations that is converted to the curvature perturbations compatible with observations 6 through the entropic mechanism. The primordial tensor perturbations are highly suppressed during the ekpyrotic contracting phase so that the tensor-to-scalar ratio is negligible and naturally agrees with the observations. The goal of the present work is to extend the viability of the two-field DHOST bouncing model constructed in 7 to the non-linear level. To this end, we refine the model in such a way that • its viability at the linear regime is not altered, • it predicts the non-Gaussianity of the primordial curvature perturbations compatible with the observations, • it does not suffer from the strong-coupling issue 48 ; 10 ; 40 ; 23 ; 22 . Observational consistency of any bouncing model at the non-linear level can be investigated by using non-Gaussianity. The main source of non-Gaussianity for the model 7 is the non-linearity of the conversion process 25 ; 49 ; 50 ; 24 . And since the modes of observational interest are super-horizon during the conversion phase, only non-Gaussianity of the local type is the relevant. If all the higher-order operators are suppressed during the conversion phase, the local non-Gaussianity is well under control, so that it can be made to agree with the observations by adjusting the model parameters 25 . However, for the model 7 higher-order operators can be relevant for the conversion process, which can amplify the non-Gaussianity such that the local non-Guassianity parameter f_{NL} may not satisfy the observational constraint -6\leq f_{NL}\leq 4.2 6 . This is one reason why we need to improve the model 7 . Another reason comes from the strong coupling issue that has to be avoided for the viability of a bouncing model at the non-linear level. Once the strong coupling is present, higher order terms in perturbation expansion of the Lagrangian are not suppressed, which may render classical description of the bouncing dynamics inconsistent and even induce non-unitarity 22 . The model of 7 has conventional asymptotics in the far past and far future so that it is free from the strong-coupling issue or high non-Gaussianities in the asymptotic regions, c.f. 1 ; 5 ; 3 ; 4 ; 2 . The strong-coupling issue can, however, arise during the bounce phase since the sound speed squared of the model becomes much smaller than 1 . In many cases the ratio of the higher-order terms in perturbation expansion of the Lagrangian to the second-order terms is inversely proportional to the sound speed squared 10 , and thus smallness of the sound speed signals the strong-coupling of the model. In the present work we show that the strong-coupling scale for our improved bouncing scenario is well above the characteristic background energy scale all the time to confirm that the strong-coupling issue does not arise. The rest of this paper is organized as follows. In Section 2 we briefly review the original two-field DHOST bouncing scenario constructed in 7 to refine such that all the higher-order operators are relevant only during the bounce phase and determine the allowed parameter space where the local non-Gaussianity agrees with the observations. In Section 3 we evaluate the strong-coupling scale of the model and show that the model does not suffer from the strong-coupling issue. We summarize the result and conclude in Section 4 . Following 33 , we adopt the signature of the metric to be (-,+,+,+) and work in reduced Planck units ( c=\hbar=1 and M_{pl}\equiv 1/\sqrt{8\pi G_{N}}=1 ) throughout this paper. 2 Two-field DHOST bounce and non-Gaussianity In this section, we scrutinize the model originally proposed in 7 closely and refine it in order to make the non-Gaussianity under control. 2.1 General setup The model studied in 7 is described by an exceptional subclass of quadratic DHOST Ia theory coupled to an extra luminal scalar field \chi . Its action is given by 7 ; 58 ; 59 \displaystyle{\cal S}_{\text{full}}={\cal S}_{\phi}+{\cal S}_{\chi}. (1) {\cal S}_{\phi} is the action of the exceptional subclass of DHOST Ia theories, namely 44 ; 12 ; 47 ; 58 ; 59 \displaystyle{\cal S}_{\phi}=\int d^{4}x\sqrt{-g}\left(F(\phi,X)+F_{2}(\phi,X)R[g]+\sum_{I=1}^{5}A_{I}(\phi,X)L_{I}\right), (2)

Read the complete text