REVIEW 3 major objections 4 minor 14 references
A sandwich with segment convexity
T0 review · 3 major / 4 minor · reviewed 2026-08-27 · deepseek-v4-flash
Pith's one-line read The paper proves that on reduced Birkhoff–Beatley systems (including Cartan–Hadamard manifolds), condition (2) together with drop completeness of the vertical extension yields a convex separator between two functions.
desk verdict Correct sufficiency theorem for convex separators, but the advertised Cartan-Hadamard application may be empty because the key drop-completeness hypothesis is never verified outside Euclidean space. 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 central object is a reduced Birkhoff–Beatley system: a set $X$ with a family of lines and a metric such that each line is metrically a copy of $\mathbb{R}$ (a ruler), which yields segments and hence a notion of convexity without vector-space structure. The load-bearing mechanism is the vertical extension $X^*=X\times\mathbb{R}$, with metric $d^*((x_0,y_0),(x_1,y_1))=\sqrt{d(x_0,x_1)^2+(y_0-y_1)^2}$; it is itself a reduced Birkhoff–Beatley system (Lemma 3). The drop-complete property is the identity $$ \operatorname{conv}(\{x_0\}\cup K)=\bigcup_{x\in K}[x_0,x] $$ for every convex $K$, and it is what makes the main induction go through. In the Cartan–Hadamard case the vertical extension is the Riemannian product with $\mathbb{R}$ (Lemma 5).
What would settle it
Independently recompute the two height coordinates $\varepsilon_1 = d(a,x)/d(a,b)$ and $\varepsilon_2 = \ln(\sqrt{17}/3)/\ln(\sqrt{41}/3)$ from the proof of Theorem 3: if the point $X_1=((0,\sqrt{17}),\varepsilon_1)$ actually lies on the union of segments from $A$ to $[B,C]$, then the claimed failure of drop completeness in $H^2\times\mathbb{R}$ is wrong and the paper's justification for its main hypothesis collapses. A direct check of Theorem 1 itself would be to find a drop-complete vertical extension with functions $f,g$ satisfying (2) but no convex separator.
Extended reading notes
Core claim
The central claim is Theorem 1: for a convex set $D$ in a reduced Birkhoff–Beatley system whose vertical extension is drop complete, the condition that for all $n$, all $x_0,\dots,x_n\in D$, all $x\in\operatorname{conv}\{x_1,\dots,x_n\}$, and all $t\in[0,1]$, $$ f(c(t))\le (1-t)g(x_0)+t f(x), $$ with $c$ the segment from $x_0$ to $x$, is sufficient for the existence of a convex separator $f\le\phi\le g$. The proof takes the convex hull of the epigraph of $g$, shows by induction that it stays above $f$, and defines $\phi$ as the infimum of the allowed heights; drop completeness of the vertical extension is exactly what lets the induction decompose the hull of a new point and a finite set into segments from that point. The paper frames this as a correction of a recent false generalization to Riemannian manifolds and as a true extension of the classical vector-space sandwich theorem.
Load-bearing premise
The load-bearing premise is that the vertical extension—the space obtained by adding a real height coordinate to every point with the product metric—has the drop-complete property, meaning that adding a point to any convex set produces exactly the union of all segments from that point to the set; if that identity fails, the induction that constructs the separator from the epigraph's convex hull breaks down.
Editorial extensions
If this is right
- When the vertical extension is drop complete, inequality (2) is a certificate that a convex separator exists, and the proof shows the separator can be taken as the lower boundary of the convex hull of the epigraph.
- On Cartan–Hadamard manifolds, Corollary 1 turns (2) into a sufficient convex-separation criterion using geodesic segments.
- If the vertical extension has convex-hull cardinality number $\kappa$, only convex hulls of at most $\kappa$ points need be checked in the hypothesis (Theorem 2).
- The hyperbolic-plane example shows drop completeness of the underlying space does not transfer to the vertical extension, so the hypothesis in the main theorem cannot simply be replaced by a condition on the original space.
- In finite-dimensional vector spaces the new condition is implied by the classical mixed-convexity inequality (1), but the reverse fails; the paper therefore gives a strictly sufficient condition in this generality, not a characterization.
Reading between the lines
- A natural testable extension is to classify which nonpositively curved symmetric spaces have drop-complete vertical extensions; the explicit three-point construction in the hyperbolic plane provides a template for such a classification.
- The paper's iterative hull algorithm could be reused as a numerical search tool: run it on products of hyperbolic spaces and test for 'lifted' convex-hull points analogous to the one in Theorem 3, which would indicate where drop completeness survives.
- Combining Theorem 2 with an explicit value of the convex-hull cardinality number for particular manifolds would turn the sufficient condition into a finite, checkable criterion; the paper leaves that computation open.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proves a conditional convex-separation theorem in the setting of reduced Birkhoff–Beatley systems, which are metric spaces with incidence and ruler axioms that include Cartan–Hadamard manifolds as examples. The main result, Theorem 1, states that if the vertical extension of the underlying system is drop complete, then a pair of functions f,g satisfying inequality (2) admits a convex separator phi with f <= phi <= g. The proof uses the convex hull of the epigraph of g and an induction over finite hulls; the drop-completeness hypothesis supplies the needed representation of convex hulls as unions of segments. Theorem 2 claims a Caratheodory-number reduction of the hypothesis, and Section 3 translates the results to Cartan–Hadamard manifolds. Section 4 constructs an example showing that the hyperbolic plane is drop complete while its vertical extension is not, and it provides a computational illustration.
Significance. If the main theorem is correct, it is a genuinely abstract sandwich theorem that avoids convex combinations entirely and applies to any reduced Birkhoff–Beatley system whose vertical extension is drop complete. This is a meaningful contribution to the literature on convexity without linear structure. The negative example in Theorem 3 is also valuable, as it shows that drop completeness of the vertical extension is a nontrivial hypothesis. However, the significance is tempered by the absence of any non-Euclidean Cartan–Hadamard example satisfying the drop-completeness hypothesis, and by several presentation gaps in the proof of Theorem 3 and in the statement of Theorem 2. The central inductive argument of Theorem 1 is sound, and the paper is honest about the strength of its assumptions.
major comments (3)
- [Section 4, Theorem 3] The proof contains a geometric inconsistency that needs to be fixed. The text says that the intersection of [p,q] and [b,c] is x=(0,sqrt(17)), but in the Poincare half-plane model the geodesic [b,c] is the semicircle X^2+Y^2=41, whose intersection with the vertical line x=0 is (0,sqrt(41)), not (0,sqrt(17)). The point (0,sqrt(17)) is instead the intersection of [p,q] with the vertical geodesic [a,r]. More importantly, the final exclusion argument is incomplete: from X1 != X2 alone it does not follow that X1 is outside the union of segments {[A,D] : D in [B,C]}. One must add the uniqueness argument that the only point D in [B,C] whose geodesic segment from A passes through x is R, because the geodesic through a and x is the vertical ray [a,r] and it meets [b,c] only at r. Please correct the typo and supply this missing justification.
- [Section 2, Theorem 2] The proof of Theorem 2 is omitted with the remark that it is essentially the same as the proof of Theorem 1. This is not entirely automatic: the induction in Theorem 1 proves the claim for hulls of arbitrary finite cardinality, and reducing to n <= kappa requires a separate argument that the Caratheodory property of the vertical extension allows the induction to stop at hulls of size kappa. Since the inductive step for a hull of m points invokes inequality (2) with n = m-1, the precise indexing should be explained. Please provide the proof or a detailed justification of the reduction.
- [Abstract and Corollary 1] The abstract promises a sufficient condition for convex separation on Cartan-Hadamard manifolds, but Corollary 1 is conditional on drop completeness of the vertical extension, and the paper gives no example of a non-Euclidean Cartan-Hadamard manifold whose vertical extension is drop complete. In fact, Theorem 3 shows that H2 x R is not drop complete, so the hypothesis is not automatic and may even be restricted to Euclidean spaces as far as the present examples show. The authors should either provide a non-Euclidean example satisfying the hypothesis or explicitly state in the abstract and introduction that the concrete non-Euclidean Cartan-Hadamard case remains open.
minor comments (4)
- [Section 1] The assertion that the generalization in reference [13] is false is made without proof or a specific counterexample. Since this is a strong claim about a published paper, the authors should substantiate it or soften the statement.
- [Section 2 and 4] There are typographical inconsistencies: 'Beatly' appears instead of 'Beatley' in Theorem 2 and in the discussion before Theorem 2, and 'Carthéodory' appears instead of 'Carathéodory' in the paragraph following Theorem 2.
- [Section 4, Theorem 3] The inequalities epsilon1 < 2/5 and epsilon2 > 2/5 are said to be checkable by hand, but no derivation is shown. A short verification of at least one of these bounds would improve readability and confidence in the example.
- [Section 4, paragraph after Theorem 3] The statement that two-dimensional Cartan-Hadamard manifolds are drop complete is asserted without proof. If this claim is needed, a proof or a precise reference should be supplied; otherwise it should be removed or explicitly marked as a conjecture.
Circularity Check
No significant circularity: Theorem 1 is a genuine sufficiency theorem proved from explicit hypotheses; no fitted inputs or self-citation chain carries the result.
full rationale
The paper's central result, Theorem 1, is a mathematical sufficiency theorem and its proof is self-contained from the stated axioms and hypotheses. The hypothesis that the vertical extension is drop complete is an explicit structural assumption, not a restatement of the conclusion; the proof uses it to decompose convex hulls in the epigraph, then applies inequality (2) to carry the induction and finally defines the separator as the lower envelope of the convex hull of the epigraph. No parameter is fitted, no empirical quantity is renamed as a prediction, and no load-bearing step reduces to a self-citation. The citation to the authors' earlier work [2] is used only for background on convexity without convex combinations and for a remark about two-dimensional Cartan-Hadamard manifolds; it is not the basis of Theorem 1 or Corollary 1. The skeptical concern that no non-Euclidean Cartan-Hadamard manifold is exhibited satisfying drop completeness is a question of applicability or hypothesis reach, not circularity: the theorem is still a valid implication from its assumptions. Similarly, the observation that conditions (2) are only sufficient, not necessary, is explicitly acknowledged and does not make the argument circular. Overall, the derivation chain is independent of its conclusion.
Assumptions & free parameters
assumptions (3)
- domain assumption Incidence postulate: any two distinct points determine a unique line.
- domain assumption Ruler postulate: each line admits a bijection with R preserving distances.
- domain assumption Drop completeness of the vertical extension.
Cite this review
Pith. "Pith review of A sandwich with segment convexity." pith.science (2026). https://pith.science/paper/E36CZTXM
@misc{pith2026200913400,
author = {Pith},
title = {Pith review of: A sandwich with segment convexity},
year = {2026},
howpublished = {\url{https://pith.science/paper/E36CZTXM}},
note = {Machine review of arXiv:2009.13400}
}
read the original abstract
The aim of this note is to give a sufficient condition for pairs of functions to have a convex separator when the underlying structure is a Cartan--Hadamard manifold, or more generally: a reduced Birkhoff--Beatley system. Some exotic behavior of convex hulls are also studied.
Reference graph
Works this paper leans on
-
[13]
A. A. Shaikh, R. P. Agarwal, and C. K. Mondal, Geodesic sandwich theorem with an application, Math. Inequal. Appl. 23 (2020) 161–167
work page 2020
- [1]
-
[2]
M. Bessenyei and B. Popovics, Convexity without convex combinations, J. Geom. 107 (2016), 77–88
work page 2016
-
[3]
G. D. Birkhoff, A set of postulates for plane geometry, based on scale and protractor, Ann. of Math., 33 (1932), 329–345
work page 1932
-
[4]
G. D. Birkhoff and R. Beatley, Basic geometry, 3rd ed., Chelsea Publishing Company, New York, 1959
work page 1959
-
[5]
C. Carathéodory, Über den Variabilitätsbereich der Fourierschen Konstanten von positiven harmonischen Funk- tionen, Rend. Circ. Mat. Palermo 32 (1911), 193–217
work page 1911
-
[6]
R. Hartshorne, Geometry: Euclid and beyond, Undergraduate Texts in Mathematics, Springer-Verlag, New York, 2000
work page 2000
-
[7]
D. Hilbert, The foundations of geometry (1899) , The Open Court Publishing Company, University of Illinois, 1950
work page 1950
Show all 14 references
-
[8]
Jost, Nonpositive curvature: geometric and analytic aspects, Lectures in Mathematics ETH Zürich, Birkhäuser Verlag, Basel, 1997
J. Jost, Nonpositive curvature: geometric and analytic aspects, Lectures in Mathematics ETH Zürich, Birkhäuser Verlag, Basel, 1997
1997
-
[9]
Kantorovitch, The method of successive approximations for functional equations, Acta Math
L. Kantorovitch, The method of successive approximations for functional equations, Acta Math. 71 (1939), 63– 97
1939
-
[10]
Y . S. Ledyaev, J. S. Treiman, and Q. J. Zhu,Helly’s intersection theorem on manifolds of nonpositive curvature, J. Convex Anal., 13 (2006), 785–798
2006
-
[11]
J. G. Ratcliffe, Foundations of hyperbolic manifolds , Graduate Texts in Mathematics, vol. 149, Second Ed., Springer, New York, 2006
2006
-
[12]
Sakai, Riemannian geometry, Translations of Mathematical Monographs, vol
T. Sakai, Riemannian geometry, Translations of Mathematical Monographs, vol. 149, American Mathematical Society, Providence, RI, 1996
1996
-
[14]
M. L. J. van de Vel, Theory of convex structures, North-Holland Mathematical Library, vol. 50, North-Holland Publishing Co., Amsterdam, 1993. INSTITUTE OF MATHEMATICS , U NIVERSITY OF DEBRECEN , H-4010 D EBRECEN , PF. 12, H UNGARY E-mail address: besse@science.unideb.hu E-mail...
1993
Reviewed August 27, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.