Pith. sign in

REVIEW 3 major objections 6 minor 23 references

Initial conditions for Starobinsky inflation in general quadratic gravity

T0 review · 3 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

Pith's one-line read General quadratic gravity still allows Starobinsky inflation without fine-tuned initial conditions.

desk verdict A plausible numerical mapping of inflation basins with a non-negligible R_ab R^ab term, but the hidden shear-derivative initial data and lack of convergence tests leave the no-fine-tuning claim under-supported. read the letter →

arxiv 2505.04805 v1 pith:D3JCJLLP submitted 2025-05-07 gr-qc astro-ph.CO

classification gr-qcastro-ph.CO PACS 04.50.Kd98.80.Cq
keywords StarobinskyinflationquadraticgravityBianchiIcosmologyinitialconditionsshearbouncefine-tuninginflationaryattractor
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper asks whether Starobinsky inflation remains a natural outcome when the $R^2$ action of Starobinsky's model is extended to general quadratic gravity by adding the invariant $R_{ab}R^{ab}$. Using a Bianchi I (spatially flat but anisotropic) cosmology with nonzero shear, the authors map which initial values of the Hubble rate $H$ and Ricci scalar $R$ produce at least 60 e-folds of inflation. They find that although the shape of this basin changes considerably with shear, successful inflation never requires fine-tuning: small shear barely disturbs the pure-$R^2$ picture, and at large shear an initially contracting band ($H<0$) bounces and then inflates. If correct, the coefficient $\alpha$ of the extra term need not be much smaller than $\beta$ for inflation to be a natural outcome, as long as $\alpha$ lies in the stability domain.

What carries the argument

The central object is the dynamical system for a diagonal Bianchi I metric (a spatially flat but anisotropic cosmology with diagonal shear) in quadratic gravity, evolved with the constraint (A.1) and the third-order equations (A.2)-(A.4), with $\sigma_-=0$ and $\sigma_+$ fixed on a grid. The authors scan a grid of initial conditions $(H,R)$, which through $R = 6(\sigma_+^2 + 2H^2 + \dot H)$ fixes $\dot H$, and classify each trajectory as giving sufficient inflation, insufficient inflation, or a Big Crunch singularity. The load-bearing mechanism is the appearance of turning points ($\dot H=0$) once shear is large enough: these can either recollapse the universe into a singularity or provide a bounce that converts an initially contracting phase into Starobinsky inflation.

What would settle it

Repeat the basin classification with an independent high-precision integrator and refine the grid in $H$ and $R$ (for instance, halving the spacing around the reported boundaries); if the band of $H<0$ initial conditions that inflate disappears or the Big Crunch outcomes near $\sigma_+ \approx 2\times10^{-6}$ shift materially, the no-fine-tuning claim would be refuted. A minimal version: take one initial condition well inside the reported $H<0$ band and verify under stricter tolerances that it bounces and reaches the quasi-de Sitter attractor.

Watch

Extended reading notes

Core claim

For the Lagrangian $L = \frac{1}{16\pi G}\left(R + (\beta-\frac{1}{3}\alpha)R^2 + \alpha R_{ab}R^{ab}\right)$ with $\beta>0$, $\alpha<0$, and the specific choice $\alpha=-10\beta$, Starobinsky inflation is stable against shear, and the set of initial conditions $(H,R)$ that yield sufficient inflation has positive measure at every shear studied. At very small shear ($\sigma_+ = 10^{-10}$) the basin is almost identical to the pure $R^2$ result; around $\sigma_+ \approx 2\times10^{-6}$ turning points ($\dot H=0$) appear, sending some trajectories to a Big Crunch singularity. For larger shear ($\sigma_+ \sim 0.1$ to $0.25$) the originally good $H>0$ region moves beyond the physically plausible rectangle $|H|<1$, $|R|<1$, but a band of good initial conditions with $H<0$ grows, and trajectories from this band bounce and then enter Starobinsky inflation. The conclusion is that no fine-tuning of initial conditions is needed for Starobinsky inflation in general quadratic gravity.

Load-bearing premise

The no-fine-tuning conclusion rests on trusting the numerical integration to classify trajectories correctly near $\dot H=0$ and near singularities, yet the paper reports no integrator, error tolerances, or grid-convergence tests.

Editorial extensions

If this is right

  • Within the stability domain, choosing $|\alpha|$ of order $\beta$ (here $\alpha=-10\beta$) does not prevent Starobinsky inflation from arising naturally from generic initial data.
  • At large shear, the successful initial conditions for inflation are predominantly initially contracting universes that bounce, rather than initially expanding ones.
  • Some initial conditions that inflate in the absence of shear instead collapse to a Big Crunch once shear exceeds roughly $2\times10^{-6}$, so shear can spoil otherwise good starts.
  • The shape of the inflation basin in the $(H,R)$ plane depends sharply on the value of shear, so the shear magnitude must be specified when discussing initial conditions for quadratic-gravity inflation.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • We infer that the same bounce mechanism could make Starobinsky inflation natural for any $\alpha$ inside the stability domain, not just $\alpha=-10\beta$; mapping the $H<0$ band for other $\alpha$ values would test this.
  • If the $H<0$ bounce band persists under refinement, estimates of the pre-inflationary probability of inflation in quadratic gravity should include shear-dominated contracting states, which could change the predicted likelihood of inflation relative to pure $R^2$.
  • The sharp appearance of Big Crunch outcomes near $\sigma_+ \approx 2\times10^{-6}$ suggests a physical threshold: an initially expanding, shear-dominated universe that inflates must have begun below this shear, otherwise it must start in the contracting bounce band.
  • A related inference for other higher-derivative gravity theories is that wherever a massive spin-2 ghost is present, shear may trigger energy transfer into the ghost and drive recollapse, so turning-point-based bounce basins may be a generic feature rather than specific to this action.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The manuscript studies the initial-condition space for Starobinsky inflation in general quadratic gravity with Lagrangian R + (β - α/3)R^2 + α R_{ab}R^{ab}, specialized to a Bianchi I (spatially flat, anisotropic) background. The authors fix α = -10β, which lies in the stability domain, and numerically integrate the system in Appendix A for several values of initial shear σ+ (with σ- = 0). They plot basin diagrams in the (H,R) plane, classifying initial data by whether they produce at least 60 e-folds of Starobinsky inflation, insufficient inflation, or a Big Crunch. The main reported results are: (i) for very small shear the basin resembles the pure R^2 case of Ref. [19]; (ii) for shear near 2×10^-6 some trajectories recollapse to a Big Crunch; (iii) for large shear (σ+ around 0.1-0.25) the H>0 inflationary region moves beyond |H|<1, but a band of initially contracting H<0 initial conditions bounces and then inflates. The authors conclude that Starobinsky inflation in general quadratic gravity does not require fine-tuning of initial conditions, and that |α| need not be much smaller than β provided α is in the stability domain.

Significance. The qualitative claim is interesting and, if confirmed, extends Refs. [19,20] in a non-trivial way: it suggests that the additional curvature invariant R_{ab}R^{ab} does not destroy the naturalness of Starobinsky inflation as long as the coefficient α lies in the stability domain. The paper is transparent about the coefficient choice and about the existence of ghost/spin-2 instabilities at large shear. Its strengths are the explicit field equations and constraint in Appendix A, the clear separation of stability regimes, and the physically motivated question. However, the numerical basin classification, which is the entire evidence for the conclusion, is currently underspecified as an initial-value problem and lacks convergence checks; for this reason the result is plausible but not yet established.

major comments (3)
  1. [Section III, Eqs. (A.1)-(A.4)] The statement that fixing σ± and choosing H and R determines all remaining initial data is incomplete. Equation (A.1) contains σ̇+, σ̈+, σ̇-, and σ̈-; unless these derivatives are specified, the constraint cannot determine Ḧ, and Eqs. (A.2)-(A.4) cannot be integrated without initial values for σ̇± and σ̈±. As written, each point in the (H,R) plane corresponds to a family of initial data with different shear derivatives, so the black/white/gray labels in Figs. 1-3 are not single-valued. The authors should state explicitly the full initial data (for instance, all first and second shear derivatives set to zero) and test whether small nonzero shear velocities or accelerations at the same (H,R) move trajectories across the basin boundaries. If the basin labels change under such perturbations, the no-fine-tuning conclusion is underdetermined.
  2. [Section III, Figs. 1-3] The paper reports no numerical integrator, tolerances, grid spacing, or convergence tests. The new qualitative features—the Big Crunch basin appearing near σ+ = 2×10^-6 and the bounce band for H<0—are inferred from trajectories that pass close to turning points (Ḣ = 0) and singularities, where numerical accuracy is most delicate. Without a resolution study, it is not possible to assess whether the basin boundaries are physical or grid-dependent, and the robustness of the central claim is therefore not established.
  3. [Section III, definition of sufficient inflation] The classification of 'sufficient' inflation uses a fixed 60-e-fold threshold, but no sensitivity test is reported. Since the basin boundaries in Figs. 1-3 are the evidence for the no-fine-tuning claim, the authors should state how the bands change when the required number of e-folds is varied (for example, 50 or 70), or at least justify that the chosen threshold does not affect the qualitative conclusion.
minor comments (6)
  1. [Section II] There is a typo in 'stable vaccum Minkowski space' (should be 'vacuum').
  2. [Section III] The sentence 'In any cases, realization of Starobinsky inflation does not need a fine-tuning...' should read 'In any case, ...'.
  3. [Section IV] The word 'bassin' appears twice in the Conclusions and should be 'basin'.
  4. [Section III] The eigenfrequency list 'λ1, λ2 = (...) H λ3, λ4 = (...) H' should be punctuated (e.g., with semicolons) to avoid ambiguity.
  5. [Section III] The restriction to the rectangle 0<R<1, -1<H<1 is justified only by a brief statement about Planck units; since the final 'always a band' claim is made for this restricted region, a slightly fuller justification or a remark on how the conclusion would change outside it would be helpful.
  6. [Fig. 3 caption] In the caption of Fig. 3, 'panel c) σ+0.2' is missing an equals sign and should read 'σ+ = 0.2'.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the numerical basin classification in (H,R) is not equivalent to any fitted input or self-citation chain.

full rationale

The paper's central claim is a numerical classification of which initial conditions in the (H,R) plane lead to at least about 60 e-folds of Starobinsky-like inflation in general quadratic gravity with fixed α=-10β and fixed shear σ±. No parameter is fitted to the target outcome: the stability domain β>0, α<0 is taken from external references (Stelle; van Dam and Veltman) and independently re-derived in the paper by linearizing around the inflationary solution. The choice α=-10β is a fixed input, not a value calibrated to produce the basin, so the no-fine-tuning conclusion is a statement about the measured extent of the basin rather than an identity. The self-citations [19,20] are used only for comparison and continuity with the authors' earlier R+R^2 studies; the shear-dependent basins, turning points, bounce region, and Big Crunch outcomes are computed in this paper, not imported as conclusions. The skeptical concern that the (H,R) plane underspecifies the fourth-order shear initial data (Section III vs. Appendix A.1-A.4) is a well-posedness and completeness issue, not a circularity: it does not make the output equal to the input by construction, and it does not involve fitting or self-referential reasoning. No step in the derivation reduces Eq. X to Eq. Y by definition, and no load-bearing premise is justified solely by a self-citation. Accordingly, the circularity score is 0, with no specific circular steps identified.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

No new particles, fields, forces, or conserved quantities are introduced; the paper only varies a known coupling alpha. The main loaded choices are the hand-picked alpha/beta ratio, the 60 e-fold threshold, the Bianchi I ansatz, the Planck-unit cutoff, and the unverified numerical accuracy.

free parameters (2)
  • alpha/beta ratio = -10
    Chosen by hand in Section II: 'we choose alpha = -10 beta', within the stability domain. The basin results and the final generalization to all stable alpha values depend on this single hand-picked ratio.
  • Required e-fold threshold = approximately 60 e-folds
    Section III: 'For appropriate initial conditions for inflation we suppose on the order of 60 e-folds.' This threshold defines which points count as sufficient inflation and therefore shapes the plotted basin area.
assumptions (5)
  • domain assumption beta > 0 and alpha < 0 ensure stability of Minkowski space and of the inflationary solution in the linearized theory.
    Quoted from Stelle and van Dam-Veltman in Section II; no derivation is repeated. The choice alpha = -10 beta and the stability claim rest on these cited results.
  • domain assumption The eigenvalue expressions lambda1 to lambda8 with negative real parts for H > 0, beta > 0, alpha < 0 correctly describe stability of the Starobinsky solution against shear.
    Section II states these frequencies after linearization; the stability domain used in the paper is inferred from them.
  • domain assumption Diagonal Bianchi I with sigma_minus = 0 and zero spatial curvature is sufficient to capture the effect of anisotropy on initial conditions.
    Section II restricts to this metric; the paper does not study nonzero sigma_minus or spatial curvature, so the conclusion is conditional on this ansatz.
  • ad hoc to paper Initial conditions outside -1 < H < 1 and 0 < R < 1 in Planck units are physically doubtful and can be excluded.
    Section III: 'going beyond 1 in Planck units is doubtful from obvious physical reasons'; this cutoff affects the claim that a good band always remains for large shear.
  • ad hoc to paper The numerical solver resolves trajectories near turning points and singularities correctly.
    No integrator, tolerances, or convergence checks are provided; the new qualitative features, such as Big Crunch outcomes and the bounce basin, depend on this assumption.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Initial conditions for Starobinsky inflation in general quadratic gravity." pith.science (2026). https://pith.science/paper/D3JCJLLP

@misc{pith2026250504805,
  author       = {Pith},
  title        = {Pith review of: Initial conditions for Starobinsky inflation in general quadratic gravity},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/D3JCJLLP}},
  note         = {Machine review of arXiv:2505.04805}
}
abstract

We consider initial conditions leading to Starobinsky inflation in the general quadratic gravity, where the action of the theory contains one more curvature square invariant in addition to $R^2$. We have chosen corresponding coefficients in a way so that the inflationary solution keeps to be stable. Our numerical results show that despite the configuration of initial conditions in the $(H,R)$ plane, leading to Starobinsky inflation can change considerably from the $R^2$ theory, realization of inflation does not need a fine-tuning of the initial conditions.

Figures

Figures reproduced from arXiv: 2505.04805 by the authors.

Figure 1
Figure 1. FIG. 1: a) Basin for very small shear [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2: For this plot the shear is increased to [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3: We present several basin plots for different initial shear. Each black point marks an initial [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 16 canonical work pages

  1. [19]

    S. S. Mishra, D. M¨ uller, and A. V. Toporensky, Phys. Rev. D 102, 063523 (2020), arXiv:1912.01654 [gr-qc]

  2. [1]

    A. A. Starobinsky, Phys. Lett. B 91, 99 (1980)

  3. [2]

    P. A. R. Ade et al. (Planck), Astron. Astrophys. 594, A20 (2016), arXiv:1502.02114 [astro-ph.CO]

  4. [3]

    Weyl, Sitzungsber

    H. Weyl, Sitzungsber. K¨ onigl. Preuss. Akad. Wiss.26, 465 (1918)

  5. [4]

    Buchdahl, Il Nuovo Cimento Series 10 23, 141 (1962)

    H. Buchdahl, Il Nuovo Cimento Series 10 23, 141 (1962)

  6. [5]

    Ruzmaikina and A

    T. Ruzmaikina and A. Ruzmaikin, Sov. Phys. JETP 30, 372 (1970)

  7. [6]

    Tomita, T

    K. Tomita, T. Azuma, and H. Nariai, Progress of Theoretical Physics 60, 403 (1978)

  8. [7]

    A. L. Berkin, Phys.Rev. D44, 1020 (1991)

Show all 23 references
  1. [8]

    Cotsakis and J

    S. Cotsakis and J. Miritzis, Class. Quant. Grav. 15, 2795 (1998), arXiv:gr-qc/9712026 [gr-qc]

  2. [9]

    Miritzis, J

    J. Miritzis, J. Math. Phys. 44, 3900 (2003), arXiv:gr-qc/0305062 [gr-qc]

  3. [10]

    J. D. Barrow and S. Hervik, Phys.Rev. D73, 023007 (2006), arXiv:gr-qc/0511127 [gr-qc]

  4. [11]

    J. D. Barrow and S. Hervik, Phys.Rev. D74, 124017 (2006), arXiv:gr-qc/0610013 [gr-qc]

  5. [12]

    Clifton and J

    T. Clifton and J. D. Barrow, Classical and Quantum Gravity 23, 2951 (2006). 8

  6. [13]

    Miritzis, Gen

    J. Miritzis, Gen. Rel. Grav. 41, 49 (2009), arXiv:0708.1396 [gr-qc]

  7. [14]

    Cotsakis and A

    S. Cotsakis and A. Tsokaros, Physics Letters B 651, 341 (2007)

  8. [15]

    Cotsakis, Gravitation and Cosmology 14, 176 (2008)

    S. Cotsakis, Gravitation and Cosmology 14, 176 (2008)

  9. [16]

    Carloni, A

    S. Carloni, A. Troisi, and P. K. S. Dunsby, Gen. Rel. Grav. 41, 1757 (2009), arXiv:0706.0452 [gr-qc]

  10. [17]

    J. D. Barrow and S. Hervik, Phys.Rev. D81, 023513 (2010), arXiv:0911.3805 [gr-qc]

  11. [18]

    Schmidt, International Journal of Geometric Methods in Modern Physics 4, 209 (2007), arXiv:gr- qc/0602017

    H.-J. Schmidt, International Journal of Geometric Methods in Modern Physics 4, 209 (2007), arXiv:gr- qc/0602017

  12. [20]

    M¨ uller and A

    D. M¨ uller and A. Toporensky, Eur. Phys. J. Plus139, 358 (2024), arXiv:2310.15377 [gr-qc]

  13. [21]

    K. S. Stelle, Gen.Rel.Grav. 9, 353 (1978)

  14. [22]

    van Dam and M

    H. van Dam and M. Veltman, Nucl.Phys. B22, 397 (1970)

  15. [23]

    Stephani, D

    H. Stephani, D. Kramer, M. MacCallum, C. Hoenselaers, and E. Herlt, Exact solutions of Einstein’s field equations (Cambridge University Press, 2003)

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.