{"id":"188ace1a-b892-4da7-a15e-43aadedbe541","arxiv_id":"2608.10871","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"In modified-gravity theories, spin Hall deflection of light and gravitational waves by a spinning black hole becomes non-commuting, so a ray disperses into a blob, and black-hole spin direction can tighten causality bounds on EFT coefficients.","lead":"This paper works out how light and gravitational waves are bent and rotated when they scatter off a spinning black hole in modified theories of gravity. It finds that the extra interactions can turn a clean split ray into a blurred blob and that the black hole's spin direction can tighten the allowed size of those extra interactions.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The numerical causality refinements in Eqs. (5.7)-(5.11) are optimized at b=b_c, where the one-loop time delay is acknowledged (footnote 17) to be potentially invalidated by higher-loop corrections; the claimed 30% and 580x enhancements are therefore not established.","rationale":"After reading the paper in good faith, I find the central noncommutative-kick calculation internally coherent: the K matrices in (4.7)-(4.8) and (4.17)-(4.18) genuinely do not commute when C_beta, C_{alpha'^2}, or C_zeta are turned on, and the figures show the joint numerical range is an ellipse. The Magnusian/eikonal identification is a practical working definition in this setup: Eq. (3.9) effectively defines the computed chi from the amplitude, and the nested-bracket terms in the generator equation compensate the known physical-versus-eikonal impact-parameter distinction at one loop. The reader's concern there is real but less decisive than the b=b_c issue. The place where the central quantitative claim is least secure is the evaluation of causality bounds at the critical impact parameter. The authors themselves flag this in footnote 17. The strongest claim as summarized gives specific numbers (30% and 580x), and those numbers are ratios of quantities at b=b_c with Gm/b ~ 1/2-1/3; the one-loop/2PM approximation is not controlled there. This is an internal limitation, not a disagreement with external consensus. I therefore recommend keeping the CONDITIONAL verdict: the paper should either soften the numerical claims, evaluate the bounds at impact parameters where the PM expansion is under better control, or provide evidence that higher-loop corrections do not shift the optimized bounds. The abstract's 'slightly' versus the 580x factor is a secondary presentation issue that should also be corrected.","tokens_in":26604,"tokens_out":11490,"duration_ms":124510,"concrete_test":"Recompute the causality bounds (5.6)-(5.10) at b = 2 b_c and b = 3 b_c for the same Kerr geodesic data, still using the one-loop time delay, and compare the optimized spin-refinement factors with the b=b_c values. If the 30% and 580x enhancements are substantially reduced or change sign as b moves away from b_c, the claimed refinements are driven by extrapolating the one-loop time delay into the uncontrolled strong-field regime. Alternatively, compute the leading 3PM (two-loop) correction to the massive-massless time delay at b=b_c and check whether it changes Delta t - Delta t_GR by an amount comparable to the EFT contributions used in the bounds.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 5 derives the spin-refined IR causality bounds by maximizing epsilon_PM^2 Z_beta and epsilon_PM^4 Z_{alpha'} over Kerr geodesics at the critical impact parameter b=b_c (Eqs. 5.8 and 5.11). At b=b_c, epsilon_PM = Gm/b is 1/2 for prograde extremal Kerr and 1/3 for non-spinning, so the post-Minkowskian expansion is not in a regime where the one-loop (2PM) results can be trusted to the claimed precision. Footnote 17 explicitly concedes that subleading PM corrections may invalidate the one-loop time delay (5.4)/(5.15) as b approaches b_c and may significantly modify the bounds. Since the headline numbers--about 30% tightening of the beta bound and about 580x tightening of the alpha'^2 bound--are ratios evaluated at exactly b=b_c for extremal Kerr, a 3PM correction of order (Gm/b)^2 ~ 0.1-0.25 relative to the retained terms can change the optimal spin orientation and the enhancement factors by O(1). The qualitative statement that spin can refine causality constraints may survive, but the quantitative claims in the abstract and in Eqs. (5.7)-(5.10) are not supported unless the b=b_c evaluation is shown to be stable.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies wave scattering on Kerr black holes in higher-derivative effective field theories of gravity using the Magnusian formalism, promoted to a matrix to track helicity. It computes gravitational Faraday rotation, wavenumber kicks (gravitational spin Hall effect), and Shapiro/Wigner–Smith time delays for photons and gravitons, at tree level and one loop, organized by post-Minkowskian, spin, and wave-optics expansions. The main claims are that EFT corrections make the parallel and orthogonal wavenumber kicks noncommuting for generic helicity states (Eq. 4.9), so a scattered ray disperses into an ellipse rather than splitting, and that black-hole spin can refine IR causality bounds on the EFT coefficients β and α'^2 by about 30% and 580x respectively (Eqs. 5.7–5.10), together with qualitative correlations among R^4 Wilson coefficients (Fig. 10). The computations are checked against published results for the eikonal phase matrix and frame-dragging.","tokens_in":26885,"tokens_out":2643,"duration_ms":25571,"significance":"If the technical assumptions are validated, the paper would provide a substantial extension of the Magnusian approach to helicity-dependent scattering observables, with a genuinely new qualitative prediction (elliptical dispersion due to noncommuting kicks) and a concrete demonstration that spin orientation can be used as a lever to sharpen IR causality constraints. The manuscript is careful in several ways: the master integrals are cross-validated using both Forde's method and LiteRed2, the time-delay and Faraday-rotation formulas are checked against refs. [2,20,21], and the final results are expressed in terms of the original Wilson coefficients without fitted parameters. The spin-refinement mechanism itself is plausible and the qualitative statement that spin tightens causality bounds may well survive; however, the quantitative enhancement factors are evaluated at the edge of the post-Minkowskian expansion and rely on an unproven identification of the Magnusian with the eikonal phase, so the headline numbers should not be taken at face value until those points are addressed.","major_comments":[{"comment":"The entire observable pipeline identifies the exact Magnusian χ with the eikonal phase δ: footnote 2 states that the two are 'numerically the same at sufficiently low orders' and that the authors 'will compute the Magnusian as the eikonal phase.' This is a load-bearing assumption because the wavenumber kicks (4.6)–(4.8), (4.17)–(4.18) and the time delays (5.3), (5.14) are extracted from χ_G1 + χ_G2 via the scattering generator equation (2.1). The asserted numerical equality at the one-loop, linear-in-spin order is not demonstrated. If the true Magnusian differs from the eikonal phase at this order, the noncommutativity (4.9) and the causality bounds of Sec. 5 would shift. The authors should either provide a direct computation of the Magnusian at this order (or a proof that the difference starts at higher order), or clearly state that the results are conditional on this identification and discuss the possible size of the correction.","section":"Sec. 2.1, footnote 2; Sec. 3.3.1"},{"comment":"The vanishing of the double-bracket term in the time delay, Eq. (5.1), is asserted without a displayed computation: 1/2! {χ_G1,{χ_G1,X_2^0}} = O(C_#^2). This term is needed to justify the simplified time-delay formula (5.2) and hence the causality bounds (5.4)–(5.10). Since nested Poisson brackets are the defining feature of the Magnusian formalism, this is not a minor omission. The authors should show the explicit computation (or provide a reference where it is carried out for these specific χ_G1 and χ_G2) rather than stating the result. If the double bracket has nonzero spin-dependent terms at O(G^2) not suppressed by the EFT couplings, the spin-refinement analysis of Sec. 5 would need to be redone.","section":"Sec. 5.1, Eq. (5.1)"},{"comment":"The claimed enhancements of the causality bounds are evaluated at the critical impact parameter b = b_c, where epsilon_PM = Gm/b is 1/2 for prograde extremal Kerr and 1/3 for non-spinning Kerr [see Eq. (5.11)]. At these values the post-Minkowskian series is not a convergent expansion, and the one-loop (2PM) result may receive O(epsilon_PM^2) ~ 0.1–0.25 corrections from 3PM terms. Footnote 17 explicitly concedes that subleading post-Minkowskian corrections may invalidate the one-loop time delay as b → b_c and may significantly modify the bounds. Since the 30% and 580x figures are ratios evaluated exactly at b = b_c, they are not established. The authors should either (i) compute or bound the 3PM corrections to the time delay, (ii) show that the optimal spin orientation and the enhancement factors are stable under O(epsilon_PM^2) perturbations, or (iii) replace the quantitative claims with the more modest qualitative statement that spin can refine causality constraints, supported by values at fixed b where the expansion is controlled.","section":"Sec. 5.1, Eqs. (5.7)–(5.11) and footnote 17"}],"minor_comments":[{"comment":"The phrase 'the gravitational spin Hall effect in EFTs of gravity are qualitatively different' has a subject-verb agreement issue ('effect... are'); please revise to 'is'.","section":"Abstract"},{"comment":"The action (3.1) mixes explicit powers of 1/Λ with the expansion of ℏ, which is discussed in Sec. 3.2, but the notation C_# is introduced before its definition. I recommend clarifying immediately after Eq. (3.1) that C_# denotes the dimensionless Wilson coefficients defined in Eq. (3.4).","section":"Sec. 3.1, Eq. (3.1)"},{"comment":"The caption of Fig. 8 refers to 'the hierarchy (3.7)' but the parameters chosen actually violate the hierarchy; the caption says this in the text, yet the first sentence could be read as claiming the opposite. Please rephrase to avoid ambiguity.","section":"Sec. 4.2.2, Figs. 8 and 9"},{"comment":"The values in Eq. (5.8) are presented as reference values for Kerr geodesics, but the text does not specify how the maximum over Euler angles (θ, ψ) is performed or whether the quoted numbers are exact or numerical. Please state the procedure and the precision.","section":"Sec. 5.1, Eq. (5.8)"},{"comment":"The footnote numbering in the sentence preceding Eq. (5.20) refers to footnote 19, but the corresponding footnote marker is missing in the main text; please check the footnote numbering throughout Sec. 5.","section":"Sec. 5.2, Eq. (5.20)"}],"recommendation":"major_revision","confidential_remarks":"The paper presents a novel and well-executed computation, and the qualitative spin-Hall noncommutativity result is likely correct and interesting. The main technical uncertainties are fixable in principle: the Magnusian versus eikonal identification, the double-bracket term, and the b = b_c stability. If the authors cannot compute the 3PM correction, they should soften the quantitative claims. The manuscript fits JHEP well, but I would not accept it in the current form because the headline numbers rest on assumptions that are explicitly acknowledged as potentially invalid in the text."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a serious paper with a new qualitative result—noncommutative wavenumber kicks from EFT corrections, Eq. (4.9)—that I do not think is in the closely related eikonal-phase-matrix paper [2]. The compute-then-diagonalize ordering is a legitimate and clearly explained contrast with ref. [2], and the resulting elliptical dispersion of the beam is a real conceptual point. The one-loop amplitude work is detailed and cross-checked (Forde and LiteRed2), the GR limits match published results, and the time delays check against ref. [2]. The spin-refined causality analysis is new in scope, and the authors are honest to present the R4 plots as “evidence” rather than as bounds.\n\nThe main load-bearing assumption is the identification of the Magnusian with the eikonal phase, acknowledged in footnote 2. It is standard at low orders, but it is not demonstrated at the order needed here; since all observables are extracted from χ_G1 + χ_G2 through the scattering generator equation, the noncommutativity in Sec. 4 and the bounds in Sec. 5 inherit this assumption. A direct check of at least the first nested bracket, or an explicit statement of where the identification could fail, would strengthen the paper considerably.\n\nThe more concrete problem is the causality analysis. Section 5 maximizes over Kerr geodesics at b=b_c, where epsilon_PM is 1/3 to 1/2 and the one-loop 2PM time delay is not in a regime where omitted subleading corrections can be neglected. Footnote 17 concedes exactly this. The headline numbers—about 30% tightening for the beta bound and about 580x for the alpha'^2 bound—are ratios evaluated at b=b_c for extremal or near-extremal Kerr, and 3PM corrections of order (Gm/b)^2 ~ 0.1–0.25 can change the optimal spin orientation and the enhancement factors by O(1). The qualitative statement that spin can refine causality constraints likely survives, but the quantitative claims in the abstract and in Eqs. (5.7)–(5.11) are not established as stated. There is also an internal mismatch: the abstract calls the enhancement “slightly,” while Eq. (5.10) quotes ~580x; that should be reconciled.\n\nMinor points: Eq. (5.1) asserts that the double-bracket term is O(C_#^2) without a displayed computation; it is probably true, but a check would help. The R4 plots are honestly framed, so no problem there. The citation pattern is fine; the self-citations are to the papers that introduced the Magnusian and the black-hole factors actually used.\n\nVerdict: this deserves a serious referee. The central novelty is real, the computations are careful, and the concerns are about interpretation of the bounds, not the core amplitude algebra. A revision that either computes or bounds the subleading PM corrections at b=b_c, or softens the quantitative causality claims to qualitative ones, would make the paper publishable. I would bring it to reading group and would cite the noncommutative kick result if it survives peer review.","headline":"A careful amplitude computation with a genuinely new qualitative effect—noncommutative wavenumber kicks—but the spin-refined causality numbers are overstated because they are evaluated at b=b_c, where the 2PM expansion is not under control.","tokens_in":27464,"tokens_out":2575,"would_cite":true,"duration_ms":26273,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"At one loop, photon and graviton wavenumber kicks in EFTs of gravity do not commute, so a spinning black hole disperses an incoming ray into an elliptical blob instead of splitting it; black-hole spin also tightens IR-causality bounds on…","keywords":["Gravitational Faraday rotation","Gravitational spin Hall effect","Magnusian matrix","Effective field theory of gravity","Scattering amplitudes","Kerr black hole","Time delay","IR causality"],"falsifier":"Compute the one-loop Magnusian directly from the modified diagrammatic rules of the Magnusian formalism (not via the eikonal identification) for the $\\beta$-corrected photon Compton amplitude, and test whether $[K^\\parallel, K^\\perp]$ and the eigenvalues of $\\partial\\chi/\\partial\\omega$ reproduce Eqs. (4.9), (5.6), and (5.9); a discrepancy would show the blob effect and the spin-refined bounds are artifacts of the identification.","tokens_in":26359,"feed_emoji":"🌀","tokens_out":10025,"duration_ms":85946,"temperature":0.7,"pith_summary":"This paper asks what a spinning black hole does to light and gravitational waves when general relativity is extended by the higher-dimension effective field theory (EFT) corrections $F F R$, $R^3$, and $R^4$, and answers with scattering-amplitude calculations. Using a helicity-matrix version of the Magnusian—the logarithm of the $S$-matrix—the authors derive the polarisation rotation angle, the wavenumber kick (deflection), and the Shapiro/Wigner–Smith time delay at one loop, to linear order in black-hole spin. The central claim is that in these EFTs the parallel and orthogonal components of the wavenumber kick do not commute, $[K^\\parallel, K^\\perp] \\neq 0$, so an incoming ray does not split into two polarisation rays as in general relativity but disperses into an elliptical blob. A second claim is that black-hole spin can be tuned to sharpen infrared-causality bounds on EFT couplings: a roughly 30% tightening of the $|\\beta|$ bound and a factor-of-about-580 tightening of the $\\alpha'^2$ bound. If true, spin orientation becomes a practical dial for constraining quantum-gravity effects, and the noncommuting kicks give a qualitatively new wave-scattering signature.","feed_headline":"Black-hole spin smears deflected light into a blob, not two rays","feed_subtitle":"EFT wavenumber kicks stop commuting, and Kerr spin tightens causality bounds on gravity couplings up to ~580x.","key_machinery":"The central object is the Magnusian matrix $\\overleftrightarrow{\\chi}$, the helicity-space generalisation of the Magnusian $\\chi=(\\hbar/i)\\log S$, with entries $\\chi_{IJ}$ for helicities $I,J=\\pm$. Observables are generated by the scattering generator equation $O_{\\rm out}=e^{\\{\\chi,\\bullet\\}}[O_{\\rm in}]$, with the wavenumber kick $k_3^\\mu=e^{\\{\\chi,\\bullet\\}}[k_2^\\mu]$ and the time delay $\\Delta t=\\partial\\chi/\\partial\\omega+\\frac12\\{\\chi,\\partial\\chi/\\partial\\omega\\}+\\cdots$; the matrix structure is needed because helicity-flipping elements make the kick and time-delay operators matrix-valued. The Magnusian matrix is obtained by Fourier-transforming tree-level and one-loop massive–massless Compton amplitudes (massive leg = Kerr black hole, massless leg = photon or graviton) to impact-parameter space, with the amplitudes built by gluing black-hole three-point factors and EFT three- and four-point vertices across $t$-channel cuts and evaluating triangle, box, and crossed-box master integrals. Holomorphic impact-parameter variables $b,\\bar b$ and the spin vector parametrisation $\\vec S=ma(\\sin\\psi\\cos\\theta,\\sin\\psi\\sin\\theta,\\cos\\psi)$ enter the Fourier kernels and determine the ellipse tilt. The noncommutativity (4.9) comes from the fact that the similarity transform diagonalising the Magnusian matrix depends on the impact parameter, so diagonalising before computing kicks and computing kicks before diagonalising are inequivalent operations.","core_discovery":"At linear order in black-hole spin and leading order in the EFT couplings, the wavenumber kick matrices in the directions parallel and orthogonal to the impact parameter fail to commute, $[K^\\parallel, K^\\perp]\\neq 0$ (Eq. 4.9), for both photon and graviton scattering. Because the two components cannot be simultaneously diagonalised, a monochromatic incoming ray does not split into two distinct, helicity-resolved rays the way a Stern–Gerlach beam does; instead, the possible polarisation states trace out a solid ellipse in wavenumber space whose centre is the unpolarised kick and whose tilt is set by spin. For photon scattering the effect is driven by the $F F R$ coupling $\\beta$ and is present even for a non-spinning black hole, which earlier eikonal treatments missed by diagonalising the phase matrix before taking derivatives. For Faraday rotation, the photon angle reproduces the general-relativity result, while the graviton angle is modified by the parity-even $R^4$ coupling and becomes wavelength-dependent. The paper also finds that black-hole spin refines the infrared-causality bounds on EFT coefficients: the $|\\beta|$ bound improves by about 30% and the $\\alpha'^2$ bound by a factor of about 580 when prograde orbits at near-critical impact parameters are used, and for $R^4$ graviton corrections the spin refinement correlates the Wilson coefficients $C_{\\zeta_1}$ and $C_{\\zeta_3}$ rather than bounding them independently.","pith_inferences":["An observational consequence: if the noncommutative kicks persist, gravitational lensing of polarised light or gravitational waves by a spinning compact object in a modified-gravity regime would show depolarisation or an elliptic smeared image rather than two clean images, which could distinguish EFT corrections from general relativity with polarimetry.","Because the tightening factors (about 30% and about 580x) come from tuning orbit orientation, observations or experiments with optimally aligned prograde, near-critical orbits would be the most sensitive probes of these couplings.","A direct one-loop calculation of the true Magnusian (rather than identifying it with the eikonal phase) would show whether the noncommutativity (4.9) survives; the authors note the identification is an assumption, so the blob effect may shift or disappear at that order.","The time-delay definition (2.9) extends to multiparticle kinematics, so a natural next step is to derive causality and positivity constraints on the $\\chi$-matrix from 3-to-3 amplitudes, going beyond 2-to-2 subprocesses."],"forward_implications":["For photon scattering with $C_\\beta\\neq 0$, the expected wavenumber kicks fill an ellipse; the ellipse collapses to a line in the general-relativity limit $C_\\beta\\to 0$, so measuring the blob-versus-split structure is a direct test of the $F F R$ coupling.","For graviton scattering, EFT corrections to the spin Hall effect are suppressed by $\\epsilon_{\\rm EFT}^4$ relative to quantum corrections (versus $\\epsilon_{\\rm EFT}^2$ for photons), so the wave-optics contribution $K^\\parallel_3$ dominates in the hierarchy-compatible regime.","Tuning the black-hole spin orientation (retrograde versus prograde orbits and adjusted impact parameters) tightens the causality bounds: roughly 30% for $|\\beta|$ from photon time delay and about $580\\times$ for $\\alpha'^2$, with smaller black holes giving stronger constraints.","The $R^4$ graviton sector shows that spin effects correlate the Wilson coefficients $C_{\\zeta_1}$ and $C_{\\zeta_3}$, turning two independent positivity conditions into a correlated condition."],"supporting_citations":[{"why":"Introduces the Magnusian and the scattering generator equation from which all observables are computed.","marker":"[1]"},{"why":"Establishes the eikonal phase matrix, deflection angle, and time delay in EFTs of gravity that this paper extends to the Magnusian matrix; Eqs. (5.3) and (5.14) are checked against its results.","marker":"[2]"},{"why":"Provides the spinning black-hole three-point and four-point amplitudes (black hole factors) used to construct the Compton amplitudes.","marker":"[19]"},{"why":"Gives the general-relativity gravitational Faraday rotation from on-shell amplitudes, the baseline for the photon rotation angle.","marker":"[20]"},{"why":"Computes quantum corrections to frame-dragging in scattering amplitudes, the baseline for graviton Faraday rotation.","marker":"[21]"},{"why":"Formulates the causality constraint that a resolvable negative time delay is forbidden, the IR condition used in Sec. 5.","marker":"[46]"},{"why":"Shows how causality constraints on gravitational EFT coefficients translate into bounds on the UV cutoff, the form adopted for the Lambda bounds.","marker":"[57]"},{"why":"Supplies Kerr geodesic bending-angle and critical-impact-parameter reference values used in the numerical bounds (5.8), (5.11), and (5.20).","marker":"[72]"},{"why":"Provides the one-loop quantum-gravity light-bending amplitude giving the b^{-3} wavenumber-kick baseline for photons.","marker":"[80]"},{"why":"Provides the one-loop quantum-gravity graviton-bending amplitude giving the baseline for graviton wavenumber kicks.","marker":"[82]"}],"fun_headline_variants":["Spin makes wavenumber kicks noncommute, smearing light rays","Gravitational spin Hall effect gets a noncommutative twist","Black-hole spin sharpens causality bounds on gravity EFTs","Deflected light smears into a blob when spin is included","Noncommuting kicks reshape gravitational spin Hall effect"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole calculation assumes the Magnusian equals the eikonal phase at the orders used, an equality the paper notes is not proven, so all kicks and time delays—and the noncommuting-kick and causality conclusions built on them—stand or fall with that identification.","fun_headline_variants_meta":{"raw":{"variants":["Spin makes wavenumber kicks noncommute, smearing light rays","Gravitational spin Hall effect gets a noncommutative twist","Black-hole spin sharpens causality bounds on gravity EFTs","Deflected light smears into a blob when spin is included","Noncommuting kicks reshape gravitational spin Hall effect"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000227,"raw_usage":{"total_tokens":1510,"prompt_tokens":1023,"completion_tokens":487,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":639,"completion_tokens_details":{"reasoning_tokens":401}},"tokens_in":639,"tokens_out":487,"duration_ms":5603,"temperature":1.0,"reasoning_tokens":401,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T15:21:51.801179+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the one-loop Magnusian directly from the modified diagrammatic rules of the Magnusian formalism (not via the eikonal identification) for the $\\beta$-corrected photon Compton amplitude, and test whether $[K^\\parallel, K^\\perp]$ and the eigenvalues of $\\partial\\chi/\\partial\\omega$ reproduce Eqs. (4.9), (5.6), and (5.9); a discrepancy would show the blob effect and the spin-refined bounds are artifacts of the identification.","supporting_citations":[],"review_version":1}