{"id":"5dff7d6d-c79f-4481-a50a-405ee6670560","arxiv_id":"1908.10663","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In Horndeski gravity, matter bounce models can simultaneously satisfy the tensor-to-scalar ratio bound and non-Gaussianity constraints, evading a no-go theorem that applies to k-essence bounce models.","lead":"These authors show that a matter bounce universe can avoid a previously established no-go theorem if the contracting phase is governed by the general scalar-tensor Horndeski theory rather than a simple k-essence field. They compute scalar and tensor non-Gaussianities and argue that tensor shapes can distinguish bounce models from inflation.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed evasion of the no-go theorem rests on the unverified assumption that perturbations pass through the bounce unchanged; without a bounce-transfer analysis for the specific Horndeski example, the observational claim remains conditional.","rationale":"The reader identified the same load-bearing assumption: the statistical nature of perturbations is assumed not to change through the bounce and expansion. This is indeed the most fragile step in the argument. The paper computes perturbation spectra and bispectra at the end of the contracting phase, but the observational bounds on r and fNL apply after the bounce. Without a demonstration that the specific Horndeski example, once completed by beyond-Horndeski operators, preserves the perturbation statistics, the central claim is conditional. I find no internal inconsistency in the perturbative calculation: the scaling assumption (7) is used consistently, the explicit example satisfies the stated conditions at a point, and the fNL/r cancellation mechanism is plausible as an existence proof. The paper's own caveat in Sec. I is phrased as an assumption, not a proven result, and the cited prior work does not cover this model. Therefore the concern is real but does not require changing the reader's conditional verdict; it reinforces it. The proposed test, computing the bounce transfer for a complete nonsingular model, would settle whether the concern actually invalidates the observational claim.","tokens_in":17082,"tokens_out":12916,"duration_ms":132237,"concrete_test":"Construct an explicit nonsingular completion of the model (73)-(81) by matching the contracting Horndeski phase to an expanding phase using healthy beyond-Horndeski operators as in Refs. [8,9], then numerically evolve the second- and third-order perturbation actions through the bounce. Compute the late-time curvature and tensor power spectra and the scalar bispectrum; compare the late-time r and fNL with the end-of-contraction values in Eqs. (33), (44), and (66)-(68). If the transfer through the bounce is k-independent and leaves r<0.064 and fNL=O(1) unchanged, the concern is resolved. If the transfer is k-dependent or changes the amplitudes by more than O(1), the claimed evasion of the observational no-go theorem is not established.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that the Horndeski example evades the no-go theorem is only observationally meaningful if the spectra and bispectra computed at the end of the contracting phase survive to late times. The paper states explicitly in Sec. I: 'Throughout the paper we assume that the statistical nature of these primordial perturbations does not change during the subsequent bouncing and expanding phases.' This assumption is load-bearing because r and fNL are evaluated at t=tb in Eqs. (33), (44), and (66)-(68); any k-dependent rescaling or mode mixing during the bounce would alter the observable r and fNL. The cited justification, Ref. [23], concerns a Horava-Lifshitz bounce and is not shown to apply to the beyond-Horndeski completion that the authors themselves say is needed to avoid gradient instabilities in the full nonsingular history. Thus the concrete Lagrangian (73)-(81) demonstrates that the no-go theorem is evaded for the contracting-phase perturbations, but not that a complete matter bounce model satisfies the observational bounds. The paper is honest about this conditional status, but the condition is not checked.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies scalar and tensor primordial perturbations during a power-law contracting phase in Horndeski gravity. Assuming a common scaling law for the background terms (Eq. (7)), the authors derive scale-invariant power spectra for curvature and tensor perturbations, compute the scalar bispectrum and the corresponding nonlinearity parameters at squeezed, equilateral and folded configurations, and compute the tensor bispectrum. They reproduce the previously known k-essence no-go result that small tensor-to-scalar ratio forces large scalar non-Gaussianity, and then construct a G2-G3 example in which r can be made small while f_NL remains O(1), thereby evading the no-go theorem during the contracting phase. They also argue that tensor non-Gaussianity from such contracting models is of squeezed type and model-dependent, in contrast to inflation, and therefore provides a possible observational discriminant.","tokens_in":17323,"tokens_out":12299,"duration_ms":130222,"significance":"If the results hold, the paper gives a concrete existence proof that the matter-bounce no-go theorem is not a generic feature of single-field scalar-tensor theories, and it identifies a potentially useful observational discriminant between contracting and inflationary scenarios. The derivation is systematic, uses established Horndeski perturbation theory, includes the full cubic action in an appendix, and connects the results to existing Planck constraints. The main caveat is that the observational conclusion rests on an explicit assumption about the bounce-to-expansion transition that is not established for the concrete Horndeski example.","major_comments":[{"comment":"The observational claim is load-bearing on the assumption stated in Sec. I that the statistical nature of the perturbations does not change during the subsequent bouncing and expanding phases. The power spectra and f_NL are evaluated at the end of the contracting phase, t=t_b, so any k-dependent rescaling or mode mixing during the bounce would alter the observable r and f_NL. The cited justification, Ref. [23], concerns a Horava-Lifshitz bounce, and no argument is given that it applies to the beyond-Horndeski completion that the authors themselves say is needed to avoid gradient instabilities in the full nonsingular history. The manuscript should either provide a bounce-transfer analysis for the Horndeski example or explicitly limit the conclusions to the contracting phase; as written, the title and abstract claim more than is demonstrated.","section":"Sec. I; Eqs. (33), (44), (66)-(68)"},{"comment":"The central derivation assumes that all background terms E_i and P_i scale as (-t)^{2 alpha} with a common exponent, and then asserts that Sigma, Theta, G_T, F_T, G_S and F_S scale as in Eq. (19). This is a nontrivial ansatz rather than a consequence of the Horndeski field equations. The subsequent spectral-index conditions in Eqs. (25)-(32) and the f_NL formulas in Eqs. (66)-(68) all depend on this scaling. The explicit example in Sec. IV B satisfies the ansatz, but the paper does not characterize which Horndeski theories admit such backgrounds for general n. The authors should either state this as a restriction on the class of models considered or prove that Eq. (19) follows from Eq. (7) and the structure of the field equations.","section":"Sec. II, Eq. (7) and Eq. (19)"}],"minor_comments":[{"comment":"The text says \"The case of alpha = -2 (nu_s = nu_t = 2/3)\" and \"alpha = 1 - 3n (nu_s = nu_t = -2/3)\"; the correct values from Eq. (25) are 3/2 and -3/2, respectively.","section":"Appendix B"},{"comment":"The expression \"O(GT) /greaterorsimilarO(mu H)\" is a LaTeX artifact and should read \"O(G_T) \\gtrsim O(mu H)\".","section":"Appendix B"},{"comment":"The assertion that functions g2(Y) and g3(Y) satisfying the local conditions (75)-(81) exist is standard, but the paper would be more self-contained if it exhibited an explicit class of smooth functions, for instance polynomials, satisfying these conditions.","section":"Sec. IV B"}],"recommendation":"major_revision","confidential_remarks":"The main issue is the bounce-transfer assumption. I would not reject the paper on this basis if the authors either add a transfer analysis for their completing theory or carefully restate the scope to contracting-phase perturbations. The scaling ansatz in Eq. (7) should also be made explicit as a class restriction rather than a general property of Horndeski contracting backgrounds."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this is a careful, useful paper that does more than the title can fully deliver. What it actually establishes is that, within the contracting phase, Horndeski theory has enough freedom to satisfy both r<0.064 and fNL=O(1), something that k-essence matter bounce cannot do. It also gives the first general Horndeski tensor bispectra for contracting backgrounds and identifies a sharp observational discriminant: contracting models produce only squeezed-type tensor non-Gaussianity, while generalized G-inflation can also produce equilateral-type. That second part is the most novel and should survive scrutiny.\n\nThe derivations are systematic and transparent. They use standard quadratic and cubic Horndeski perturbation theory, write out the full Lambda_i expressions in the appendices, and correctly reproduce the k-essence limits from the earlier no-go literature. The concrete G2/G3 example is a legitimate existence proof: conditions (75)-(81) are local constraints on g2(Y) and g3(Y) at one point, and nothing in them looks impossible. I take that as real progress, not a sleight of hand.\n\nThe soft spots are where the claim meets observation. The paper states explicitly in Sec. I that it assumes the statistical nature of the perturbations does not change through the subsequent bouncing and expanding phases. That assumption is load-bearing, because r and fNL are evaluated at t=tb. The cited justification is a Horava-Lifshitz computation, not a Horndeski or beyond-Horndeski bounce, so the transfer issue is not settled for the example in this paper. The authors are honest about the conditionality, but the title's \"evading the no-go theorem\" should be read as \"evading it for the contracting phase until a complete nonsingular history is checked.\"\n\nTwo secondary caveats. First, Eq. (7) assumes every background term scales as (-t)^{2alpha}, and the later scaling of GT, FS, GS, FT inherits that assumption. This is a modeling assumption rather than a theorem for all Horndeski backgrounds, so the generality is not as wide as the setup suggests. Second, Appendix B notes that for the scale-invariant growing-mode case, anisotropies grow and must be assumed not to ruin the bounce before it occurs. That is another external condition, not a solved problem.\n\nBottom line: this paper deserves serious peer review and is worth reading for the tensor bispectrum shapes and the Horndeski extension of the k-essence no-go. I would cite it as a contracting-phase result, not as evidence that bounce models are observationally viable, until the transfer through the bounce is analyzed.","headline":"A careful Horndeski extension of the bounce no-go analysis with a clean tensor-bispectrum discriminant, but the observational claim of evading the no-go theorem is conditional on unproven bounce transfer.","tokens_in":17794,"tokens_out":3313,"would_cite":true,"duration_ms":37391,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["98.80.Cq","04.50.Kd"],"model":"deepseek-v4-flash","headline":"A no-go theorem for matter bounce cosmology does not hold in the more general Horndeski theory.","keywords":["primordial non-Gaussianity","matter bounce cosmology","Horndeski theory","tensor-to-scalar ratio","tensor non-Gaussianity","contracting universe","no-go theorem","cosmological perturbations"],"falsifier":"Start from the paper's concrete Lagrangian example, complete it with an explicit non-singular bounce, and evolve the perturbations through the bounce to late times; if the tensor-to-scalar ratio is small but the late-time non-Gaussianity parameter exceeds $O(1)$, or if the tensor bispectrum develops an equilateral peak, the claimed evasion of the no-go theorem fails.","tokens_in":16869,"feed_emoji":"🌌","tokens_out":17952,"duration_ms":136155,"temperature":0.7,"pith_summary":"The paper shows that a no-go theorem previously derived for k-essence matter bounce cosmology, which forbids having both a small tensor-to-scalar ratio and small non-Gaussianity of the curvature perturbation, does not hold in the more general Horndeski scalar-tensor framework. Working on a general power-law contracting background $a\\propto(-t)^n$ with $0<n<1$, the authors compute the power spectra and bispectra of scalar and tensor perturbations and exhibit a concrete Horndeski Lagrangian with $r\\ll1$ and $f_{\\rm NL}\\sim O(1)$. They further find that tensor non-Gaussianity from contracting universes is squeezed in shape and model-dependent in amplitude, unlike the inflationary predictions in which an equilateral shape is possible and the squeezed amplitude is fixed. If correct, matter bounce models remain observationally viable and can potentially be distinguished from inflation using tensor non-Gaussianity.","feed_headline":"Bounce universes evade a non-Gaussianity no-go theorem","feed_subtitle":"In general scalar-tensor theories, a bounce can have small non-Gaussianity along with a small tensor-to-scalar ratio.","key_machinery":"The central machinery is the Horndeski action (the most general second-order scalar-tensor theory) combined with the power-law scaling assumption of Eq. (7), that every background term behaves as $(-t)^{2\\alpha}$. This scaling makes the coefficients in the quadratic actions for the curvature perturbation $\\zeta$ and tensor perturbation $h_{ij}$, namely $G_S,F_S,G_T,F_T$, all proportional to $(-t)^{2(\\alpha+1)}$ and hence constant up to a common factor; it selects the scale-invariant contracting branch $\\alpha=1-3n$ and guarantees that the dimensionless coefficients $\\Lambda_i$ in the cubic scalar Lagrangian are time independent. The scalar bispectrum is computed after a field redefinition that removes the boundary term $F(\\zeta)E_S$, whose effect is encoded in the coefficients $A$ and $B$ defined from $\\Theta$ and $G_T$; the tensor bispectrum comes from the two interaction terms $\\dot{h}^3$ (with coefficient $\\mu=-\\frac{1}{2}\\partial G_T/\\partial H$) and $h^2\\partial^2h$. The concrete evasion is achieved by a Lagrangian with $G_2(\\varphi,X)$ and $G_3(\\varphi,X)$ depending on $Y=Xe^{2\\varphi/\\mu}$ that satisfies five algebraic conditions at $Y=\\bar{Y}$, leaving free small parameters $\\delta_1,\\delta_2,\\delta_3$ that control $F_S$, $G_S$, and $\\Lambda_1$ independently.","core_discovery":"The central discovery is that the incompatibility between the observational upper bound on the tensor-to-scalar ratio and the observational upper bound on scalar non-Gaussianity, which excludes matter bounce models built from a k-essence scalar, disappears when the contracting phase is described by the Horndeski theory, the most general second-order scalar-tensor theory. On a power-law contracting background $a=(-t)^n$ with $0<n<1$, the authors impose the scaling behavior $E_i,P_i\\sim(-t)^{2\\alpha}$ and find scale-invariant power spectra for curvature and tensor perturbations on the branch $\\alpha=1-3n$, where superhorizon modes grow as $|\\eta|^{-3}$ instead of freezing. Using the in-in formalism, they compute the full scalar bispectrum (with coefficients $\\Lambda_i$ made constant by the scaling) and the tensor bispectrum, which receives a new $\\dot{h}^3$ contribution from $G_{5X}\\neq0$ and a GR-type $h\\partial^2h$ contribution. They then construct a concrete example with $G_2=M_{\\rm Pl}^2\\mu^2e^{-2\\varphi/\\mu}g_2(Y)$, $G_3=M_{\\rm Pl}^2\\mu g_3(Y)$, $G_4=M_{\\rm Pl}^2/2$, and $G_5=0$, and tune the functions so that $F_S\\simeq\\frac{3}{5}\\delta_1 M_{\\rm Pl}^2$ and $G_S\\simeq\\frac{3}{5}\\delta_2 M_{\\rm Pl}^2$, giving $r=16\\delta_1^{3/2}\\delta_2^{-1/2}\\ll1$ and $c_s^2=\\delta_1/\\delta_2=O(1)$ while the dangerous coefficient $\\Lambda_1$ is suppressed by a third small number $\\delta_3$; the result is $f_{\\rm NL}\\lesssim1$. For tensor modes, both the new and GR interactions yield bispectra peaked at the squeezed limit with model-dependent amplitudes, whereas generalized G-inflation also allows an equilateral peak and has a model-independent squeezed amplitude.","pith_inferences":["A direct test of the weakest assumption would be to complete the model with an explicit beyond-Horndeski bounce and evolve the bispectrum through it; if the bounce amplifies $f_{\\rm NL}$ or changes the tensor bispectrum shape, the claimed evasion would not survive in a full non-singular history.","The construction requires the Lagrangian functions $g_2(Y)$ and $g_3(Y)$ to satisfy tuned conditions on their first three derivatives at a single point, which raises a potential fine-tuning concern; quantifying the measure of the parameter space satisfying all five conditions would clarify how generic the evasion is.","The conformal equivalence to non-attractor inflation implies that the tensor bispectrum alone may not differentiate a bounce from non-attractor inflation; adding scalar non-Gaussianity or other observables would be needed to break the degeneracy.","One could search for squeezed tensor non-Gaussianity in CMB B-mode bispectra as a direct test: a null detection of equilateral tensor non-Gaussianity together with a squeezed amplitude inconsistent with inflation's fixed value would favor the contracting scenario."],"forward_implications":["If the central claim is right, matter bounce models in Horndeski gravity are not excluded by the current bounds $r<0.064$ and $f_{\\rm NL}\\lesssim O(1)$, so the no-go theorem does not close the door on bounce alternatives to inflation.","The tensor bispectrum offers a distinguishing observable: contracting models predict only squeezed-shape tensor non-Gaussianity, so a detected equilateral peak in the tensor bispectrum would rule out this class of bounce models.","Because the squeezed tensor non-Gaussianity amplitude from contracting models depends on the Horndeski functions while the inflationary one is fixed, measuring that amplitude can distinguish the two scenarios.","Slightly detuning the relation $\\alpha=1-3n$ produces a red spectral tilt $n_s\\simeq0.96$, consistent with Planck, so the mechanism can accommodate the observed scalar tilt.","The computed tensor bispectra agree with those of non-attractor inflation models, because both are conformally equivalent to the matter-dominated contracting scenario; this implies a shared tensor signature between the two scenarios."],"supporting_citations":[{"why":"Supplies the extended no-go theorem for k-essence matter bounce, with formulas for the tensor-to-scalar ratio and non-Gaussianity that the paper reproduces and contravenes.","marker":"[21]"},{"why":"Defines the Horndeski action used throughout, establishing the framework as the most general second-order scalar-tensor theory.","marker":"[22]"},{"why":"Provides the quadratic actions and the definitions of the coefficients G_S, F_S, G_T, F_T, as well as the generalized G-inflation tensor bispectra used for comparison.","marker":"[25]"},{"why":"Gives the cubic Lagrangian for the curvature perturbation in Horndeski theory, from which the scalar bispectrum and the non-Gaussianity parameter are derived.","marker":"[28–30]"},{"why":"Supplies the tensor interaction Hamiltonian and the inflationary tensor bispectrum amplitudes used for shape and amplitude comparison.","marker":"[36]"},{"why":"Cited as justification that in some matter bounce cases the statistical nature of perturbations is unchanged through the bounce, supporting the paper's central assumption.","marker":"[23]"},{"why":"Sets the observational upper bound on the tensor-to-scalar ratio (r < 0.064) used to define what counts as small.","marker":"[26]"},{"why":"Sets the observational constraints on local and equilateral non-Gaussianity used to define what counts as small f_NL.","marker":"[35]"},{"why":"Source of the matter-bounce GR tensor bispectrum whose scale dependence appears in the GR-type amplitude.","marker":"[37]"},{"why":"Provides the non-attractor inflation squeezed tensor non-Gaussianity that the paper shows agrees with its contracting-model amplitudes by conformal equivalence.","marker":"[39]"}],"fun_headline_variants":["Horndeski bounce dodges non-Gaussianity no-go","Scalar-tensor bounce fits r and non-Gaussianity bounds","Bounce cosmology evades no-go via Horndeski","Tensor bispectra distinguish bounce from inflation"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the statistical nature of the primordial perturbations does not change during the subsequent bouncing and expanding phases, so that the spectra and bispectra evaluated at the end of the contracting phase are the ones observed.","fun_headline_variants_meta":{"raw":{"variants":["Horndeski bounce dodges non-Gaussianity no-go","Scalar-tensor bounce fits r and non-Gaussianity bounds","Bounce cosmology evades no-go via Horndeski","Tensor bispectra distinguish bounce from inflation"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000856,"raw_usage":{"total_tokens":3795,"prompt_tokens":1099,"completion_tokens":2696,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":715,"completion_tokens_details":{"reasoning_tokens":2627}},"tokens_in":715,"tokens_out":2696,"duration_ms":21615,"temperature":1.0,"reasoning_tokens":2627,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:10:56.800189+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Start from the paper's concrete Lagrangian example, complete it with an explicit non-singular bounce, and evolve the perturbations through the bounce to late times; if the tensor-to-scalar ratio is small but the late-time non-Gaussianity parameter exceeds $O(1)$, or if the tensor bispectrum develops an equilateral peak, the claimed evasion of the no-go theorem fails.","supporting_citations":[{"cited_title":"Evolution of cosmological perturbations and the production of non-Gaussianities through a nonsingular bounce: Indications for a no-go theorem in single field matter bounce cosmologies","cited_arxiv_id":"1508.04141","evidence_quote":"Supplies the extended no-go theorem for k-essence matter bounce, with formulas for the tensor-to-scalar ratio and non-Gaussianity that the paper reproduces and contravenes."},{"cited_title":"Viable tensor-to-scalar ratio in a symmetric matter bounce","cited_arxiv_id":"1703.10061","evidence_quote":"Defines the Horndeski action used throughout, establishing the framework as the most general second-order scalar-tensor theory."},{"cited_title":"Second-order scalar-tensor field equ a- tions in a four-dimensional space,","cited_arxiv_id":null,"evidence_quote":"Provides the quadratic actions and the definitions of the coefficients G_S, F_S, G_T, F_T, as well as the generalized G-inflation tensor bispectra used for comparison."},{"cited_title":"Fluctuations in a Ho\\v{r}ava-Lifshitz Bouncing Cosmology","cited_arxiv_id":"0911.3196","evidence_quote":"Cited as justification that in some matter bounce cases the statistical nature of perturbations is unchanged through the bounce, supporting the paper's central assumption."},{"cited_title":"Primordial non-Gaussianities of gravitational waves in the most general single-field inflation model","cited_arxiv_id":"1108.3513","evidence_quote":"Provides the non-attractor inflation squeezed tensor non-Gaussianity that the paper shows agrees with its contracting-model amplitudes by conformal equivalence."}],"review_version":1}