REVIEW 2 major objections 4 minor 47 references
This paper proves that mixed volumes of a convex body with its own reflection are bounded by the simplex value, settling a 1938 conjecture, and gives the sharp L_p analogue with simplex extremizers.
Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →
Godbersen's 1938 conjecture is proved: V(K[k],−K[n−k]) ≤ C(n,k) vol(K) for all convex bodies, with equality characterizations, and it yields the sharp L_p Rogers–Shephard inequality.
T0 review reviewed 2026-08-01 challenge →
load-bearing objection The inequality part of Godbersen is real and elegant; the polytope equality proof has a fixable but real gap in Lemma 4.4. the 2 major comments →
Godbersen's conjecture and the $L_p$-Rogers-Shephard inequality
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
Core claim
On its own terms, the paper establishes two theorems. Theorem 1.1: for any convex body K in R^n and 0 ≤ k ≤ n, V(K[k], -K[n-k]) ≤ (n choose k) vol(K); if K is a polytope with nonempty interior and 0 < k < n, equality holds exactly when K is a simplex. Theorem 1.2: if 0 is in K and p ∈ (1, ∞] with 1/p + 1/q = 1, then vol(K +_p (-K)) ≤ sum_{k=0}^n (n choose k)^2 (n/q choose k/q)^{-1} vol(K), with equality exactly when K is a simplex with a vertex at the origin. These are sharp refinements of the Rogers-Shephard bound, and the equality statements identify the simplex as the unique extremal shape in the relevant settings.
What carries the argument
The central device is a factorization identity for mixed volumes in complementary orthogonal subspaces of R^{2n}: for bodies L_i in E and M_j in E^⊥, the full mixed volume factors as (N choose k)^{-1} times the product of the two lower-dimensional mixed volumes, with constants normalized by Lebesgue measure. Applied to the diagonal embedding Δ(x)=(x,x) and the anti-diagonal embedding Δ̃(x)=(x,-x), mixed-volume monotonicity (ΔK ⊂ K^2) yields the binomial bound in Proposition 3.2. Equality for polytopes is then read from support conditions of the mixed surface-area measure, which force face incidence relations that only a simplex can satisfy. For the L_p result, the Firey L_p sum is represente
Load-bearing premise
The equality half of the L_p theorem rests, in the final step of Theorem 1.2, on a cited but not re-proved result that V(K, -K, ..., -K) = n vol(K) forces K to be a simplex; if that k=1 characterization had other extremizers, the equality classification in Theorem 1.2 would not follow.
What would settle it
Compute V(K[2], -K[2]) for a convex body in R^4 and compare with 6 vol(K): a value above 6 would refute the inequality, and a non-simplex polytope attaining 6 would refute the equality classification.
If this is right
- For every 0 < λ < 1, the volume ratio of (1-λ)K + λ(-K) is maximized by simplices; this confirms a separate conjecture that was previously known only in low dimensions.
- The equality result for polytopes is the first full solution of the sharp form of the conjecture in dimensions n ≥ 4, where earlier results covered only special classes such as bodies of constant width and anti-blocking bodies.
- The L_p Rogers-Shephard inequality now holds for all convex bodies containing the origin and all p > 1; previously it was known only in the plane or for restricted classes, and equality forces a simplex with a vertex at the origin.
- The method also proves the inequality part of the higher-order generalization of Godbersen's conjecture and gives a new proof of Schneider's higher-order inequality.
- Quantitative stability of the Rogers-Shephard inequality reduces to stability of the k=1 mixed-volume inequality and hence to the Minkowski measure of symmetry.
Where Pith is reading between the lines
- Because the equality classification in Theorem 1.1 is proved only for polytopes, the question of whether non-simplex convex bodies can attain equality remains open; continuity of mixed volumes does not by itself transfer the polytope result, since an approximating sequence need not attain equality.
- The p = ∞ case of Theorem 1.2, namely vol(conv(K ∪ -K)) ≤ 2^n vol(K), is a concrete specialization that could be tested by computation for high-dimensional polytopes; its equality case is a simplex with a vertex at the origin.
- The factorization argument is not obviously limited to the pair K, -K; a natural test is whether the same monotonicity yields a sharp bound for V(K[k], L[n-k]) when L is contained in a scaled copy of K, which would extend the conjecture to nonsymmetric pairs.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proves Godbersen's conjecture: for every convex body K⊂R^n and every 0≤k≤n, the mixed volume V(K[k],−K[n−k]) is at most binom(n,k) vol_n(K). For polytopes with nonempty interior and 0<k<n, it further characterizes equality by K being a simplex. The proof of the inequality is a short argument in R^{2n} using the diagonal embedding, the mixed-volume splitting lemma, and monotonicity of mixed volumes. The equality case for polytopes is obtained by converting the equality chain into vanishing of certain integrals η_{r,j}, then using face incidence conditions to force simpliciality and finally a simplex. In the last section, the paper derives an L_p Rogers–Shephard inequality for bodies containing the origin, with sharp constant Σ (n choose k)^2 (n/q choose k/q)^{-1} vol_n(K), and equality for simplices with a vertex at the origin.
Significance. If the proofs are completed, this settles a conjecture from 1938 and provides a clean, unified derivation of both the classical Rogers–Shephard inequality and its L_p analogue. The inequality part (Proposition 3.2) is elegant and appears correct; the derivation of the L_p inequality by integrating the discrete inequality is conceptually attractive. The equality classifications are substantial and are the main vulnerability of the paper: they depend on delicate face-incidence arguments and on an external characterization of equality for k=1. The manuscript also explicitly claims that the method extends to higher-order settings, which increases its potential impact. However, the equality proof as written contains a false cone identity and an induction gap, so the paper is not yet in acceptable form.
major comments (2)
- [§4, Lemma 4.4 (Eq. (15))] The proof asserts C=(ι1C1+ι2C2)∩(∆C0). This is false: the definition of C only requires v1+v2∈C0, whereas ∆C0 imposes v1=v2. Consequently, u∈relint C need not lie in relint(∆C0), and the span formula (15) is not justified as written. Since (15) is used to conclude w∈spanC, Lemma 4.4 is unproved. This lemma is load-bearing for the polytope equality case of Theorem 1.1. The likely repair is to replace ∆C0 by δ^{-1}(C0)={(v1,v2): v1+v2∈C0} and to justify (15) via relative interiors of the two cones; such a correction must be supplied.
- [§4, proof of Theorem 1.1 (induction step)] The induction step is stated for all 1≤k≤n−1, but the induction hypothesis applies in dimension n−1 only to nontrivial indices 1≤m≤n−2. For k=n−1, Lemma 4.6 propagates the vanishing of η only to j'=n−1, which is the trivial index for a facet E, so it does not imply that E is a simplex. Thus the conclusion that K is simplicial is not established for k=n−1. The gap is fixable by using the symmetry V(K[k],−K[n−k])=V((−K)[n−k],K[k]) to reduce equality for k=n−1 to the case k=1, but this reduction is not stated in the manuscript.
minor comments (4)
- [§4, Lemma 4.3] The claim 'F1∩F2=∅ implies f0≤n−1' is false: if F0=K, then u1+u2=0, not necessarily u=0; for a cube with u=(e1,−e1), opposite facets satisfy the displayed dimension identities and f0=n. This assertion is not actually needed, since r≥0 already follows from the choice s≥n, but the passage should be corrected.
- [§5, Lemma 5.1] In the definition of a1,a2,b1,b2, the factors (1−t)^{-1/q} and t^{-1/q} are undefined at t=0 or t=1. This can be handled by assuming 0<t<1 and using continuity (or by treating the endpoints separately), but it should be stated.
- [§4, Lemma 4.7] In the proof, the set A_i is used but not defined; it should presumably be V(E_i). Also, the sentence 'Then for i≠j, A_i∩A_j is an (n−2)-dimensional face' needs clarification.
- [General] Some notation is introduced informally (e.g., K^2 for K×K, and the binomial notation with real arguments). A brief note fixing these conventions would improve readability.
Circularity Check
No significant circularity: the main inequalities are derived from standard mixed-volume monotonicity with exact constants; self-citations are contextual only.
full rationale
The main inequality (5) is not taken as input: Proposition 3.2 obtains it by comparing V(K^2[n], ι1K[k], ι2K[n−k]) with V(∆K[n], ι1K[k], ι2K[n−k]) through monotonicity of mixed volumes, Lemma 2.1, and the exact identities (9)–(11). The target quantity appears only after the comparison, not in the hypothesis, and the constant C(n,k) is computed exactly, not fitted. Proposition 3.1 is the same independent argument for Rogers–Shephard. Theorem 1.2 follows by Fubini integration of the already proved inequality (5); the coefficient Σ_k C(n,k)^2 (n/q choose k/q)^{-1} is obtained from an explicit beta integral. No fitted parameter is later renamed as a prediction. The equality analysis for polytopes is internal (Lemmas 4.2–4.7) apart from the classical k=1 simplex characterization cited to Grünbaum [19, §6.1] and the simplex verification in [35,15] for the L_p case; those are independent external facts, not author-supplied inputs chosen to force the conclusion. Self-citations [24] and [17] are used only for context and motivation; the proof does not rely on them. A reviewer-flagged correctness concern in Lemma 4.4 (the asserted identity C=(ι1C1+ι2C2)∩(∆C0) and the span formula (15)) would be a proof gap if it stands, but it is not a circular reduction of the theorem to its assumptions. Thus the circularity score is low.
Axiom & Free-Parameter Ledger
axioms (6)
- standard math Mixed-volume splitting formula: (N choose k)V(L_1,...,L_k,M_1,...,M_{N-k}) = V^E(L_1,...,L_k)V^{E⊥}(M_1|E⊥,...,M_{N-k}|E⊥) for L_i⊂E, M_j⊂E⊥ (Lemma 2.1, from [44, Thm 5.3.1]).
- standard math Positivity and support characterization of mixed volumes: V(K_1,...,K_n)>0 iff dim(Σ_{i∈I}K_i) ≥ |I| for all I; support formula (7) for mixed surface area measure.
- standard math Monotonicity of mixed volume: K⊂L implies V(K,K_2,...,K_n) ≤ V(L,K_2,...,K_n).
- domain assumption Characterization of equality in the k=1 case: V(K,−K[n−1]) = n vol_n(K) iff K is a simplex (Minkowski measure of symmetry; Grünbaum [19, §6.1]).
- standard math The L_p-Minkowski sum representation K+_p L = { (1-t)^{1/q} x + t^{1/q} y : x∈K,y∈L,t∈[0,1] } for p>1, q=p/(p-1) (Lutwak–Yang–Zhang [33, Lemma 2]).
- domain assumption Simplices with a vertex at the origin attain equality in the L_p-Rogers–Shephard inequality (6) (Bianchini–Colesanti [7]; Manui–Ndiaye–Zvavitch [35, Lemma 23]; Fradelizi–Manui–Meyer–Ndiaye [15, Thm 1]).
Cite this review
Pith. "Pith review of Godbersen's conjecture and the $L_p$-Rogers-Shephard inequality." pith.science (2026). https://pith.science/paper/ZJE6DXQC
@misc{pith2026260720387,
author = {Pith},
title = {Pith review of: Godbersen's conjecture and the $L_p$-Rogers-Shephard inequality},
year = {2026},
howpublished = {\url{https://pith.science/paper/ZJE6DXQC}},
note = {Machine review of arXiv:2607.20387}
}
abstract
We prove that the mixed volume of a convex body with its reflection about the origin is maximized by simplices. This confirms a conjecture of C. Godbersen from 1938 and refines the Rogers-Shephard inequality. We also prove that, among convex polytopes, simplices are the only extremizers. Finally, we use this inequality to prove the $L_p$-version of the Rogers-Shephard inequality for convex bodies containing the origin and show that, for any $p\in(1,\infty]$, the only extremizers are simplices with a vertex at the origin.
Reference graph
Works this paper leans on
-
[1]
Alonso-Guti´ errez, S
D. Alonso-Guti´ errez, S. Artstein-Avidan, B. Gonz´ alez Merino, C. H. Jim´ enez, and R. Villa,Rogers-Shephard and local Loomis-Whitney type inequalities, Math. Ann.374(2019), no. 3-4, 1719–1771
2019
-
[2]
Alonso-Guti´ errez, B
D. Alonso-Guti´ errez, B. Gonz´ alez Merino, C. H. Jim´ enez, and R. Villa,Rogers-Shephard inequality for log- concave functions, J. Funct. Anal.271(2016), no. 11, 3269–3299
2016
-
[3]
Alonso-Guti´ errez, M
D. Alonso-Guti´ errez, M. A. Hern´ andez Cifre, M. Roysdon, J. Yepes Nicol´ as, and A. Zvavitch,On Rogers- Shephard type inequalities for general measures, Int. Math. Res. Not. IMRN10(2021), 7224–7261
2021
-
[4]
Artstein-Avidan, K
S. Artstein-Avidan, K. Einhorn, D. I. Florentin, and Y. Ostrover,On Godbersen’s conjecture, Geom. Dedicata 178(2015), 337–350
2015
-
[5]
Artstein-Avidan and E
S. Artstein-Avidan and E. Putterman,On unbalanced difference bodies and Godbersen’s conjecture, Proc. Amer. Math. Soc.153(2025), no. 12, 5345–5359. 14
2025
-
[6]
Artstein-Avidan, S
S. Artstein-Avidan, S. Sadovsky, and R. Sanyal,Geometric inequalities for anti-blocking bodies, Commun. Contemp. Math.25(2023), no. 3, Paper No. 2150113, 30
2023
-
[7]
Bianchini and A
C. Bianchini and A. Colesanti,A sharp Rogers and Shephard inequality for thep-difference body of planar convex bodies, Proc. Amer. Math. Soc.136(2008), no. 7, 2575–2582
2008
-
[8]
B¨ or¨ oczky Jr.,Around the Rogers-Shepard inequality, Math
K. B¨ or¨ oczky Jr.,Around the Rogers-Shepard inequality, Math. Pannon.7(1996), no. 1, 113–130
1996
-
[9]
Math.190 (2005), no
,The stability of the Rogers-Shephard inequality and of some related inequalities, Adv. Math.190 (2005), no. 1, 47–76
2005
-
[10]
Yu. ˜D. Burago and V. A. Zalgaller,Geometric inequalities, Translated from the Russian by A. B. Sosinski ˘i, Springer Series in Soviet Mathematics, Springer-Verlag, Berlin, 1988
1988
-
[11]
G. D. Chakerian,Inequalities for the difference body of a convex body, Proc. Amer. Math. Soc.18(1967), 879–884
1967
-
[12]
Colesanti,Functional inequalities related to the Rogers-Shephard inequality, Mathematika53(2006), no
A. Colesanti,Functional inequalities related to the Rogers-Shephard inequality, Mathematika53(2006), no. 1, 81–101
2006
-
[13]
F´ ary and L
I. F´ ary and L. R´ edei,Der zentralsymmetrische Kern und die zentralsymmetrische H¨ ulle von konvexen K¨ orpern, Math. Ann.122(1950), 205–220
1950
-
[14]
Wm. J. Firey,p-means of convex bodies, Math. Scand.10(1962), 17–24
1962
-
[15]
M. Fradelizi, A. Manui, M. Meyer, and C. S. Ndiaye,Equality cases for theL p-Rogers–Shephard inequality in the plane and for locally anti-blocking bodies inR n, preprint (2026),arXiv:2606.07887
Pith/arXiv arXiv 2026
-
[16]
,L p-Rogers–Shephard type inequalities forL p-zonoids and symmetric bodies, preprint (2026), arXiv:2607.03582
Pith/arXiv arXiv 2026
-
[17]
Fryˇ s,Godbersen’s conjecture and its generalization, Ms.C
F. Fryˇ s,Godbersen’s conjecture and its generalization, Ms.C. Thesis, Charles University, Prague, 2026
2026
-
[18]
Godbersen,Der Satz vom Vektorbereich in R¨ aumen beliebiger Dimensionen, Ph.D
C. Godbersen,Der Satz vom Vektorbereich in R¨ aumen beliebiger Dimensionen, Ph.D. Thesis, G¨ ottingen, 1938
1938
-
[19]
Gr¨ unbaum,Measures of symmetry for convex sets, Proc
B. Gr¨ unbaum,Measures of symmetry for convex sets, Proc. Sympos. Pure Math., Vol. VII, pp. 233–270, Amer. Math. Soc., Providence, RI, 1963
1963
-
[20]
Gr¨ unbaum,Convex polytopes, Prepared and with a preface by Volker Kaibel, Victor Klee and G¨ unter M
B. Gr¨ unbaum,Convex polytopes, Prepared and with a preface by Volker Kaibel, Victor Klee and G¨ unter M. Ziegler, Springer-Verlag, New York, 2003
2003
-
[21]
Haddad, D
J. Haddad, D. Langharst, E. Putterman, M. Roysdon, and D. Ye,Affine isoperimetric inequalities for higher- order projection and centroid bodies, Math. Ann.393(2025), no. 1, 1073–1121
2025
-
[22]
,Higher-orderL p isoperimetric and Sobolev inequalities, J. Funct. Anal.288(2025), no. 2, Paper No. 110722, 45
2025
-
[23]
J. Haddad, D. Langharst, G. V. Livshyts, and E. Putterman,On the polar of Schneider’s difference body, preprint (2025),arXiv:2503.06191
Pith/arXiv arXiv 2025
-
[24]
Kotrbat´ y,On a generalization of Godbersen’s conjecture, Int
J. Kotrbat´ y,On a generalization of Godbersen’s conjecture, Int. Math. Res. Not. IMRN22(2025), Paper No. rnaf341, 11
2025
-
[25]
Kotrbat´ y and T
J. Kotrbat´ y and T. Wannerer,On mixed Hodge-Riemann relations for translation-invariant valuations and Aleksandrov-Fenchel inequalities, Commun. Contemp. Math.24(2022), no. 7, Paper No. 2150049, 24
2022
-
[26]
Langharst,Some comments on the mth-order projection bodies, J
D. Langharst,Some comments on the mth-order projection bodies, J. Convex Anal. (2026), to appear
2026
-
[27]
Langharst, E
D. Langharst, E. Putterman, M. Roysdon, and D. Ye,On themth-order weighted projection body operator and related inequalities, Pure Appl. Funct. Anal.10(2025), no. 5, 1323–1353
2025
-
[28]
Langharst, M
D. Langharst, M. Roysdon, and Y. Zhao,On themth-order affine P´ olya-Szeg¨ o principle, J. Geom. Anal.35 (2025), no. 7, Paper No. 205, 32
2025
-
[29]
D. Langharst, F. M. Sola, and J. Ulivelli,Higher-order reverse isoperimetric inequalities for log-concave functions, preprint (2024),arXiv:2403.05712
arXiv 2024
-
[30]
Langharst and D
D. Langharst and D. Xi,General higher orderL p mean zonoids, Proc. Amer. Math. Soc.152(2024), no. 12, 5299–5311
2024
-
[31]
Lutwak,The Brunn-Minkowski-Firey theory
E. Lutwak,The Brunn-Minkowski-Firey theory. I. Mixed volumes and the Minkowski problem, J. Differential Geom.38(1993), no. 1, 131–150
1993
-
[32]
,The Brunn-Minkowski-Firey theory. II. Affine and geominimal surface areas, Adv. Math.118(1996), no. 2, 244–294
1996
-
[33]
Lutwak, D
E. Lutwak, D. Yang, and G. Zhang,The Brunn-Minkowski-Firey inequality for nonconvex sets, Adv. in Appl. Math.48(2012), no. 2, 407–413
2012
-
[34]
Makai Jr.,Research problems, Period
E. Makai Jr.,Research problems, Period. Math. Hungar.5(1974), no. 4, 353–354
1974
-
[35]
Manui, C
A. Manui, C. S. Ndiaye, and A. Zvavitch,On the volume of sums of anti-blocking bodies, Commun. Contemp. Math. (2026), to appear
2026
-
[36]
C. A. Rogers and G. C. Shephard,The difference body of a convex body, Arch. Math. (Basel)8(1957), 220–233
1957
-
[37]
London Math
,Convex bodies associated with a given convex body, J. London Math. Soc.33(1958), 270–281
1958
-
[38]
Roysdon,Rogers-Shephard type inequalities for sections, J
M. Roysdon,Rogers-Shephard type inequalities for sections, J. Math. Anal. Appl.487(2020), no. 1, 123958, 22
2020
-
[39]
Sadovsky,Godbersen’s conjecture for locally anti-blocking bodies, Adv
S. Sadovsky,Godbersen’s conjecture for locally anti-blocking bodies, Adv. Geom.25(2025), no. 3, 307–315
2025
-
[40]
Schneider,Eine Verallgemeinerung des Differenzenk¨ orpers, Monatsh
R. Schneider,Eine Verallgemeinerung des Differenzenk¨ orpers, Monatsh. Math.74(1970), 258–272. 15
1970
-
[41]
132– 141, Ann
,On the Aleksandrov-Fenchel inequality, Discrete geometry and convexity (New York, 1982), pp. 132– 141, Ann. New York Acad. Sci.440, New York Acad. Sci., New York, 1985
1982
-
[42]
527–53824, 2000
,Mixed functionals of convex bodies, pp. 527–53824, 2000. The Branko Gr¨ unbaum birthday issue
2000
-
[43]
Geom.96(2009), no
,Stability for some extremal properties of the simplex, J. Geom.96(2009), no. 1-2, 135–148
2009
-
[44]
,Convex bodies: the Brunn-Minkowski theory, Cambridge University Press, Cambridge, 2014
2014
-
[45]
R. van Handel and S. Wang,On Minkowski’s monotonicity problem, preprint (2025),arXiv:2507.20082
Pith/arXiv arXiv 2025
-
[46]
X. Zhou, D. Ye, and Z. Zhang,Themth order Orlicz projection bodies, J. Geom. Anal.35(2025), no. 9, Paper No. 259, 35
2025
-
[47]
G. M. Ziegler,Lectures on polytopes, Springer-Verlag, New York, 1995. Charles University, Faculty of Mathematics and Physics, Mathematical Institute of Charles University, Sokolovsk´a 49/83, 186 00 Prague, Czechia Email address:kotrbaty@karlin.mff.cuni.cz Email address:mohamed-abdeldjalil.mouamine@matfyz.cuni.cz 16
1995
This paper was first reviewed by deepseek-v4-flash on August 1, 2026.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.