REVIEW 3 major objections 6 minor 28 references
Collinear Corrections to the Cachazo-Strominger Soft Theorem
T0 review · 3 major / 6 minor · reviewed 2026-08-16 · deepseek-v4-flash
Pith's one-line read The paper establishes that the sub-subleading soft graviton theorem must include collinear correction terms.
desk verdict The distributional-correction mechanism is sound and the five-point MHV check is solid, but the universal formula (6.2) drops the {1,2} pair via an unjustified b>2 restriction, so the all-tree-level claim fails as stated. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing mechanism is a distributional identity for spinor derivatives, $$\tilde\lambda_i \partial_{\tilde\lambda_j}\frac{1}{\langle jk\rangle} = [ik]\,\delta(j,k),\qquad \delta(j,k)=\pi\$delta^{2}$(\langle jk\rangle),$$ which the paper treats as a non-vanishing distributional relation even where earlier derivations set it to zero. Applied to the product of angle-bracket poles in an MHV amplitude, the identity turns derivatives of the soft operator into delta-function terms concentrated on collinear pairs; these recombine, through the collinear factorization of the amplitude, into the graviton splitting functions times a lower-point amplitude. The central move is the prescription that these distributional terms be subtracted from the definition of the soft factor, yielding an identity in which both sides are ordinary functions.
What would settle it
Take an explicit five-point MHV graviton amplitude, apply $\frac{1}{2}\sum_a [sa]\langle sa\rangle^{-1}D_a^2$ to the four-point amplitude, and collect the coefficient of each collinear delta function $\delta(a,b)$ after imposing momentum conservation. If that coefficient is not $\frac{1}{2}[sb]^3\langle sb\rangle^{-1} f^{h_a h_b}_{h_P}(t)$ times the appropriate three-point amplitude, the corrected soft theorem (1.6) fails; conversely, a direct canonical charge computation yielding a different coefficient of the collinear delta term after the four $\bar z$-derivatives of (7.3) would falsify the claimed match.
Extended reading notes
Core claim
On the paper's terms, the correct positive-helicity sub-subleading soft graviton theorem for tree-level amplitudes is $$$M^{{(2)}}$_{n+1} = \frac{1}{2}\sum_{a=1}^n \frac{[sa]}{\langle sa\rangle} $D_a^{2}$ M_n - \frac{1}{2}\sum_{a=1}^n \sum_{\substack{b>2\\ b>a}} \frac{[sb]^3}{\langle sb\rangle} $f^{{h_a h_b}}$_{h_P}(t)\, \delta(a,b)\, M_{n-1}(\ldots, $P^{{h_P}}$, \ldots),$$ where $D_a = \tilde\lambda_s \partial_{\tilde\lambda_a}$ and the second sum subtracts the distributional collinear terms generated when $D_a^2$ acts on the amplitude's poles. The subtraction uses the graviton splitting functions $f^{h_a h_b}_{h_P}(t)$ with $t = \omega_a/(\omega_a+\omega_b)$ and delta functions $\delta(a,b)$ forcing particles $a$ and $b$ collinear. The paper derives this for all tree-level MHV amplitudes and argues by collinear factorization that it extends to general tree-level amplitudes, then shows that after translating to celestial-sphere coordinates and taking four $\bar z$ derivatives, the correction matches the non-divergent collinear component of the canonical asymptotic charge.
Load-bearing premise
The argument stands or falls on treating the distributional spinor identity as physically operative, so that derivatives acting on angle-bracket poles produce delta functions that must be subtracted; if the original convention of dropping these terms were the correct one, every collinear correction would vanish and the agreement with the asymptotic charge analysis would disappear.
Editorial extensions
If this is right
- The sub-subleading soft theorem for tree-level gravitons becomes an identity between distribution-free quantities only after the collinear subtraction; comparisons that omit it are comparing the soft factor to a shifted amplitude.
- At leading and subleading order the correction vanishes, so the Weinberg and subleading soft theorems are unaffected; the new terms are specific to sub-subleading order in gravity.
- The correction matches, after four celestial-sphere derivatives, the non-divergent collinear part of the canonical asymptotic charge, so the soft-theorem/Ward-identity correspondence survives at this order.
- For general tree-level amplitudes the same subtraction is dictated by collinear factorization, so the result is not an artifact of the MHV form used in the proof.
- By the same argument, acting with the soft factor on the collinear pole is equivalent to taking the soft limit after the collinear limit, resolving the order-of-limits ambiguity for these terms.
Reading between the lines
- Editorial inference: the subtraction rule effectively redefines the soft operator on amplitudes; checking that the shifted operator reproduces the celestial OPE algebra would give an independent test of the split between hard and collinear charge.
- Editorial inference: because the derivation relies only on the distributional identity and collinear factorization, the same construction should produce subleading-order collinear corrections in Yang-Mills theory; the paper names this as a future direction, but it follows from the same mechanism.
- Editorial inference: at loop level, where subleading soft factors are corrected, the same distributional terms are expected to appear shifted; self-dual gravity, where loop amplitudes are known, is a natural place to look for them.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. This paper argues that the sub-subleading soft graviton theorem of Cachazo and Strominger must be corrected by collinear/distributional terms. Starting from the identity (3.3), the authors note that spinor derivatives acting on angle-bracket poles of amplitudes generate delta-function-supported terms, and they propose to subtract these from the soft factor. The general corrected formula is Eq. (1.6)/(6.2). The paper verifies the prescription in detail for the five-point MHV amplitude, gives a general MHV recombination argument in Section 5, argues heuristically in Section 6 that the result extends to all tree-level amplitudes, and identifies the correction with the collinear part of the asymptotic charge found in [18].
Significance. If the corrected theorem is valid, this is a significant result: it provides an amplitude-side derivation of the collinear charge component of [18], clarifies the status of the Cachazo-Strominger soft theorem at sub-subleading order, and gives a concrete example where distributional terms in the soft expansion cannot be dropped. The five-point MHV example is a genuine, fully worked check, including the exact ϵ=1 sum in Appendix B and comparison with [21]. The comparison with [18] is an external consistency check, not a circular use of the target result, and no parameters are fitted. At the same time, the universal claim in Eq. (6.2) currently has a load-bearing summation problem and the all-tree-level extension is only sketched, so the significance would be fully realized only after those issues are resolved.
major comments (3)
- [§6.2, Eq. (6.2) (and Eq. (1.6))] The correction sum is restricted by 'b>2; b>a', so the only unordered pair excluded is {1,2}. This restriction is not derived anywhere: the general MHV formula (5.9) sums over all k≠a with no b>2 condition, and the general subtraction formula (6.1) also sums over all k≠a. The five-point MHV example used to motivate (4.12) has no 1/<12> pole, so it cannot justify dropping δ(1,2) terms. For a generic tree amplitude with a 1/<12> pole, D^2_1 (or D^2_2) acting on that pole produces a nonzero distributional term, and omitting it makes the right-hand side of (6.2) depend on which leg is labeled 1; interchanging labels 1 and 3 changes the set of subtracted pairs while the left-hand side transforms covariantly. The stress-test concern therefore lands: the universal corrected theorem as stated is not permutation invariant. The authors must either prove that all δ(1,2) terms vanish for arbitrary helicities, or symmetrize the correction sum over all unordered pairs and explain the relation between the ordered sum in (5.9) and the restricted sum in (6.2). This issue also affects the consistency check in Section 7, which relies on (6.2).
- [§5.2, Eq. (5.9)] The recombination of distributional terms in the general MHV case is asserted rather than demonstrated. Eq. (5.8) contains single delta functions, derivatives of delta functions, and products of delta functions; the text states that the products vanish and the derivative terms combine with the single-delta terms, but the general-n computation is not shown. Because this recombination is the only derivation of the correction term for MHV amplitudes, it should be presented in full, or replaced by a cleaner argument, before it is imported into Eq. (6.2).
- [§6, 'General Tree-level Graviton Amplitudes'] The extension from MHV to all tree-level amplitudes is an expectation argument, not a proof. The text says 'one expects the result to carry over' and appeals to collinear factorization, but the soft derivative D^2_a acts before the collinear limit is taken, and the commutativity of these operations is precisely what needs to be established. If the theorem is claimed for all tree-level graviton amplitudes, a derivation for non-MHV helicity configurations must be supplied, or the claim should be restricted to MHV amplitudes.
minor comments (6)
- [Footnote 5] The text first sets ∂_{\tilde\lambda_a}\langle ab\rangle^{-1}=0 in going from (2.9) to (2.10) and then reintroduces it as a distributional term in Section 3; this should be acknowledged explicitly as a choice of how the compact form is defined, to avoid the appearance of inconsistency.
- [Eq. (4.11)] The two-line computation of D^2_a([ab]/\langle ab\rangle) omits the intermediate derivative-of-delta terms that the general discussion says must be kept; expanding one line or adding a sentence explaining the cancellation would make the example easier to follow.
- [Section 7, Eq. (7.3)] The transition from (7.2) to (7.3) is very compressed, especially the replacement of the summand by f^{h_a h_b}_{h_P}(t)δ(s,b)δ(a,b) after four derivatives; a schematic derivation or a precise reference to the corresponding step in [18] should be given.
- [Appendix C] The notation ∂δ(a,b) is used without definition; it should be specified as the derivative of the delta function with respect to the relevant holomorphic coordinate.
- [Eq. (2.13)] Only three helicity combinations of the graviton splitting functions are listed; the (a^-,b^-) combination relevant for the possible δ(1,2) terms in repeated-negative-helicity MHV configurations is not given, so the claimed vanishing of those terms in that sector cannot be checked from the text.
- [Section 2, Eqs. (2.8)-(2.10)] The symbol n denotes both the number of hard particles and the reference leg in the BCFW sum; this double use is confusing in Section 5, where n is the hard multiplicity while the reference leg is 4 or 5.
Circularity Check
No significant circularity: the corrected soft theorem is derived from a standard distributional identity and checked against an independent asymptotic-charge benchmark, with no fitted parameters and no load-bearing self-citations.
full rationale
The paper's central claim is the corrected sub-subleading soft graviton theorem, Eq. (1.6)/(6.2), with a collinear subtraction term. The derivation chain is self-contained: (i) write the MHV amplitude in the pole form (5.1) via Hodges' formula; (ii) apply the standard distributional identity (3.3), λ_i ∂_{λ̃_j}(1/<jk>) = [ik]δ(j,k), to compute D_a and D_a^2 acting on the amplitude, Eqs. (5.2)–(5.3); (iii) expand the distributional terms, show subleading terms vanish by the Schouten identity, and simplify the sub-subleading terms to Eq. (5.9); (iv) extend to general tree-level amplitudes by collinear factorization in Sec. 6; and (v) compare with the independent asymptotic-charge analysis of Ref. [18] in Sec. 7. No parameter is fitted, no equation is imported from the target result, and no load-bearing premise is justified by a self-citation: the present authors do not overlap with Ref. [18] or Ref. [21]. The subtraction prescription is a definitional choice—one demands that both sides of the soft expansion be free of distributional singularities—and the distributional terms are computed, not assumed. The consistency with [18] is therefore a genuine cross-check rather than an input. The only noteworthy concern is a potential correctness gap: Eq. (6.2) restricts the correction sum to pairs with b>2 and b>a, whereas the general derivation in Eq. (5.9) sums over all k≠a, so for amplitudes with a 1/<12> pole the stated universal formula may omit a δ(1,2) term and break relabeling invariance. That is a mathematical-completeness issue, not a circularity: it concerns the validity of the claimed result rather than the result being assumed as an input.
Assumptions & free parameters
assumptions (5)
- ad hoc to paper The spinor derivative identity lambda_i tilde_d_{lambda_j} (1/<jk>) = [ik] delta(j,k), with delta(j,k) = pi delta^2(<jk>), applies during the soft expansion, and the resulting delta terms are subtracted from the soft factor.
- domain assumption The holomorphic soft limit, lambda_s -> epsilon lambda_s, tilde_lambda_s -> tilde_lambda_s, implements the soft expansion.
- domain assumption Tree-level graviton amplitudes factorize in the collinear limit with the splitting functions of Eq. (2.13).
- domain assumption The MHV amplitude can be written as M_n = F_n product_{i<j} 1/<ij> with F_n pole-free, and F_n -> [ak] F_{n-1} in the collinear limit.
- domain assumption The BCFW-based soft expansion of Eqs. (2.8)-(2.10) holds with an arbitrary reference spinor lambda_n.
Cite this review
Pith. "Pith review of Collinear Corrections to the Cachazo-Strominger Soft Theorem." pith.science (2026). https://pith.science/paper/PZLHN7OW
@misc{pith2026250416903,
author = {Pith},
title = {Pith review of: Collinear Corrections to the Cachazo-Strominger Soft Theorem},
year = {2026},
howpublished = {\url{https://pith.science/paper/PZLHN7OW}},
note = {Machine review of arXiv:2504.16903}
}
read the original abstract
Soft theorems describe the behavior of scattering amplitudes when one or several external particles are taken to be energetically soft. In tree-level gravity there are universal soft theorems for the three leading orders in the soft expansion, and they can be shown to be equivalent to Ward identities of asymptotic symmetries. While the leading and subleading symmetries are understood as supertranslations and superrotations respectively, the precise symmetry interpretation of the sub-subleading soft theorem is still a matter of investigation. The form of the sub-subleading soft graviton theorem was elucidated by Cachazo and Strominger using a BCFW expansion of graviton amplitudes. In this work we show that consistency with results based on asymptotic charges requires a careful treatment of collinear singularities in the amplitude, giving rise to collinear corrections to the usual Cachazo-Strominger soft theorem.
Reference graph
Works this paper leans on
-
[18]
Sub-sub leading soft graviton theorem from asymptotic Einstein’s equations,
L. Freidel, D. Pranzetti, and A.-M. Raclariu, “Sub-sub leading soft graviton theorem from asymptotic Einstein’s equations,” JHEP 05 (2022) 186, arXiv:2111.15607 [hep-th]
arXiv 2022
-
[21]
Evidence for a New Soft Gr aviton Theorem,
F. Cachazo and A. Strominger, “Evidence for a New Soft Gr aviton Theorem,” arXiv:1404.4091 [hep-th]
-
[1]
Infrared photons and gravitons,
S. Weinberg, “Infrared photons and gravitons,” Phys. Rev. 140 (1965) B516–B524
1965
-
[2]
M ultileg one loop gravity amplitudes from gauge theory,
Z. Bern, L. J. Dixon, M. Perelstein, and J. S. Rozowsky, “M ultileg one loop gravity amplitudes from gauge theory,” Nucl. Phys. B 546 (1999) 423–479, arXiv:hep-th/9811140
arXiv 1999
-
[3]
Tree level amplitudes from soft theorems,
K. Zhou, “Tree level amplitudes from soft theorems,” JHEP 03 (2023) 021, arXiv:2212.12892 [hep-th]
arXiv 2023
-
[4]
Infinite Set of Soft Theorems in Gau ge-Gravity Theories as Ward-Takahashi Identities,
Y. Hamada and G. Shiu, “Infinite Set of Soft Theorems in Gau ge-Gravity Theories as Ward-Takahashi Identities,” Phys. Rev. Lett. 120 no. 20, (2018) 201601, arXiv:1801.05528 [hep-th] . 20
arXiv 2018
-
[5]
Strominger, Lectures on the Infrared Structure of Gravity and Gauge Theory
A. Strominger, Lectures on the Infrared Structure of Gravity and Gauge Theory . Princeton University Press, 2018. arXiv:1703.05448 [hep-th]
arXiv 2018
-
[6]
A. Strominger, “ w1+∞ Algebra and the Celestial Sphere: Infinite Towers of Soft Gra viton, Photon, and Gluon Symmetries,” Phys. Rev. Lett. 127 no. 22, (2021) 221601
work page 2021
Show all 28 references
-
[7]
Flat Space Amplitudes and Conformal Symmetry of the Celestial Sphere,
S. Pasterski, S.-H. Shao, and A. Strominger, “Flat Space Amplitudes and Conformal Symmetry of the Celestial Sphere,” Phys. Rev. D96 no. 6, (2017) 065026, arXiv:1701.00049 [hep-th]
2017 arXiv
-
[8]
Conformally Soft Photons and Gravitons,
L. Donnay, A. Puhm, and A. Strominger, “Conformally Soft Photons and Gravitons,” JHEP 01 (2019) 184, arXiv:1810.05219 [hep-th]
2019 arXiv
-
[9]
Conformall y Soft Theorem in Gauge Theory,
M. Pate, A.-M. Raclariu, and A. Strominger, “Conformall y Soft Theorem in Gauge Theory,” Phys. Rev. D 100 no. 8, (2019) 085017, arXiv:1904.10831 [hep-th]
2019 arXiv
-
[10]
Conformally Soft Theorem in Gravity,
A. Puhm, “Conformally Soft Theorem in Gravity,” arXiv:1905.09799 [hep-th]
1905 arXiv
-
[11]
Lectures on Celestial Holography,
A.-M. Raclariu, “Lectures on Celestial Holography,” arXiv:2107.02075 [hep-th]
-
[12]
Lectures on celestial amplitudes,
S. Pasterski, “Lectures on celestial amplitudes,” Eur. Phys. J. C 81 no. 12, (2021) 1062, arXiv:2108.04801 [hep-th]
2021 arXiv
-
[13]
Hol ographic Symmetry Algebras for Gauge Theory and Gravity,
A. Guevara, E. Himwich, M. Pate, and A. Strominger, “Hol ographic Symmetry Algebras for Gauge Theory and Gravity,” arXiv:2103.03961 [hep-th]
-
[14]
Soft algebras for leaf amplitudes,
W. Melton, A. Sharma, and A. Strominger, “Soft algebras for leaf amplitudes,” JHEP 07 (2024) 070, arXiv:2402.04150 [hep-th]
2024 arXiv
-
[15]
Currents in celestial CFT,
A. Ball, “Currents in celestial CFT,” Mod. Phys. Lett. A 39 no. 29n30, (2024) 2430007, arXiv:2407.13558 [hep-th]
2024 arXiv
-
[16]
Asymptotic higher spin symme tries I: covariant wedge algebra in gravity,
N. Cresto and L. Freidel, “Asymptotic higher spin symme tries I: covariant wedge algebra in gravity,” Lett. Math. Phys. 115 no. 2, (2025) 39, arXiv:2409.12178 [hep-th]
2025 arXiv
-
[17]
Asymptotic Higher Spin Symmetries III: Noe ther Realization in Yang-Mills Theory,
N. Cresto, “Asymptotic Higher Spin Symmetries III: Noe ther Realization in Yang-Mills Theory,” arXiv:2501.08856 [hep-th]
-
[19]
Higher s pin dynamics in gravity and w1+ ∞ celestial symmetries,
L. Freidel, D. Pranzetti, and A.-M. Raclariu, “Higher s pin dynamics in gravity and w1+ ∞ celestial symmetries,” Phys. Rev. D 106 no. 8, (2022) 086013, arXiv:2112.15573 [hep-th]
2022 arXiv
-
[20]
On infini te symmetry algebras in Yang-Mills theory,
L. Freidel, D. Pranzetti, and A.-M. Raclariu, “On infini te symmetry algebras in Yang-Mills theory,” JHEP 12 (2023) 009, arXiv:2306.02373 [hep-th]
2023 arXiv
-
[22]
Subleading Soft Graviton Theorem for Loop Ampl itudes,
A. Sen, “Subleading Soft Graviton Theorem for Loop Ampl itudes,” JHEP 11 (2017) 123, arXiv:1703.00024 [hep-th] . 21
2017 arXiv
-
[23]
A brief introduction to modern amplitude m ethods,
L. J. Dixon, “A brief introduction to modern amplitude m ethods,” in Theoretical Advanced Study Institute in Elementary Particle Physics: Particle Ph ysics: The Higgs Boson and Beyond, pp. 31–67. 2014. arXiv:1310.5353 [hep-ph]
2014 arXiv
-
[24]
Celestial Operator Products of Gluons and Gravitons,
M. Pate, A.-M. Raclariu, A. Strominger, and E. Y. Yuan, “ Celestial Operator Products of Gluons and Gravitons,” arXiv:1910.07424 [hep-th]
1910 arXiv
-
[25]
A simple formula for gravitational MHV ampl itudes,
A. Hodges, “A simple formula for gravitational MHV ampl itudes,” arXiv:1204.1930 [hep-th]
1930 arXiv
-
[26]
Di Francesco, P
P. Di Francesco, P. Mathieu, and D. S´ en´ echal,Conformal field theory . Graduate texts in contemporary physics. Springer, New York, NY, 1997. https://cds.cern.ch/record/639405
1997
-
[27]
On Loop Corrections to Subleading Soft Behavior of Gluons and Gravitons,
Z. Bern, S. Davies, and J. Nohle, “On Loop Corrections to Subleading Soft Behavior of Gluons and Gravitons,” Phys. Rev. D 90 no. 8, (2014) 085015, arXiv:1405.1015 [hep-th]
2014 arXiv
-
[28]
One loop n point helicity amplitudes in (selfdual) gravity,
Z. Bern, L. J. Dixon, M. Perelstein, and J. S. Rozowsky, “ One loop n point helicity amplitudes in (selfdual) gravity,” Phys. Lett. B 444 (1998) 273–283, arXiv:hep-th/9809160. 22
1998 arXiv
Reviewed August 16, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.