Ideal points, directed completion and the case of the maximally extended Schwarzschild spacetime

Argam Ohanyan; Marco Picerni; Matteo Zanardini; Nicola Gigli; Zhe-Feng Xu

REVIEW 2 major objections 4 minor 23 references

Ideal points, directed completion and the case of the maximally extended Schwarzschild spacetime

T0 review · 2 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper proves that directed completion and future causal completion coincide for globally hyperbolic spacetimes, and describes the directed completion of the maximally extended Schwarzschild spacetime in terms of radial null geodesics.

desk verdict Section 3 is solid, but the Schwarzschild application rests on a false directed-sup-closedness claim in Lemma 4.5, so the advertised characterization is likely wrong. read the letter →

arxiv 2608.04718 v1 pith:CJGBCMLB submitted 2026-08-05 math.DG gr-qc

Nicola Gigli , Argam Ohanyan , Marco Picerni , Zhe-Feng Xu , Matteo Zanardini This is my paper · ORCID

classification math.DGgr-qc MSC 53C5006A06

keywords directedcompletionfuturecausalLorentzianpre-lengthspacesgloballyhyperbolicspacetimesSchwarzschildspacetimeradialnullgeodesicsidealpointsboundary

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 proves that two seemingly different ways of adding ideal points to a spacetime give the same result. On one side is the future causal completion, the classical construction based on indecomposable past sets; on the other is the directed completion, a purely order-theoretic construction that adjoins suprema of directed subsets. The paper shows these coincide for smooth globally hyperbolic spacetimes and, more generally, for approximable past-distinguishing Lorentzian pre-length spaces satisfying a supremum-compatibility condition. In the concrete case of the maximally extended Schwarzschild spacetime it then describes the completion explicitly, with the boundary built from radial null geodesics.

What carries the argument

The load-bearing construction is the directed completion itself: for a partially ordered set one takes the family of all lower and directed-sup-closed subsets generated by directed subsets, ordered by inclusion, with suprema given by unions. The bridge to causal geometry is the observation that every indecomposable past set is directed, together with the supremum-compatibility condition, which says that the past of a supremum is the union of the pasts, so that the embedding $x \mapsto I^-(x)$ has the monotone convergence property required by the universal property. In the Schwarzschild section the work is done by the radial null geodesics $\gamma_{U,\Omega}(s)=(U,s,\Omega)$ and the symmetric family with roles of $U$ and $V$ exchanged: angular-variation estimates along future-directed causal curves force the angular coordinates of any directed set to converge, so every directed set can be identified with the past of a single radial null geodesic.

What would settle it

A directed set in the maximally extended Schwarzschild spacetime with $U_*\in(-\infty,0)$ and $V_*=+\infty$ whose angular coordinates fail to converge—for instance, a sequence of causal segments accumulating angular oscillation at future null infinity—would violate the angular estimate (4.19) and break Theorem 4.1; looking for such a set is a direct check of the claim.

Watch

Extended reading notes

Core claim

The central claim is that the directed completion of a Lorentzian pre-length space—the partial order obtained by freely adjoining suprema of directed sets—coincides with the future causal completion, the classical construction that adds ideal points described by terminal indecomposable past sets. The coincidence holds for approximable, past-distinguishing spaces satisfying the paper's supremum-compatibility condition, and in particular for all smooth globally hyperbolic spacetimes. For the maximally extended Schwarzschild spacetime, the paper further identifies the completed poset explicitly: every directed set is, up to the closure operation, the causal past of a radial null geodesic, and the boundary elements are the two copies of future null infinity, the two future timelike infinities, and the black hole singularity, ordered by inclusion of these pasts.

Load-bearing premise

The whole construction stands on the condition that no ideal point has a least upper bound inside the spacetime, together with the angular-variation estimates in the Schwarzschild case; if either gives way, the identified completions diverge from the classical one.

Editorial extensions

If this is right

The future causal completion, not just the directed one, is the canonical directed-complete poset attached to any smooth globally hyperbolic spacetime, so the two boundary theories are the same on that class.
For every directed set with a supremum in such a spacetime, the chronological past of the supremum equals the union of the chronological pasts of its elements; equivalently, no terminal indecomposable past set has a least upper bound inside the spacetime.
The completion of the maximally extended Schwarzschild spacetime is explicitly describable: future null infinity has two copies parametrized by $(U,\Omega)$ and $(V,\Omega)$, future timelike infinity consists of two points, and the black hole singularity is parametrized by $(U,\Omega)$, with order given by inclusion of causal pasts of radial null geodesics.
Every directed set in the Schwarzschild spacetime is carried, up to closure, by a single radial null geodesic, so radial null geodesics generate the entire boundary.
The description reproduces the previously known causal completion of the Schwarzschild spacetime from the ideal-point construction, confirming that the two constructions agree in this concrete example.

Reading between the lines

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

A natural test extension is to run the same case analysis for spherically symmetric or static black-hole spacetimes with two monotone coordinates; wherever angular-variation estimates hold, the completion should again be generated by radial null geodesics.
Because the directed completion depends only on the partial order, the equivalence may be expected to hold for globally hyperbolic Lorentzian length spaces and metric spacetimes, not only smooth manifolds; the paper notes this in a remark but does not develop the applications.
The precise failure mode of the supremum-compatibility condition—an ideal point with a least upper bound—suggests that in spacetimes with naked singularities or holes, the directed completion and the future causal completion can genuinely differ; constructing such an example and comparing the two posets would sharpen the boundary between the two notions.
The angular-convergence estimates indicate a general criterion: a directed set can be assigned a unique ideal point exactly when its monotone coordinates converge and causality forces its angular coordinates to converge; this could serve as a computational shortcut for other spacetimes.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Dear colleague,

Read this one for Section 3, not for Section 4. The equivalence between future causal completion and directed completion under (SC) is a clean, rigorous result. The (SC) condition is well motivated, the proof of Theorem 3.15 is complete, and Proposition 3.17 correctly covers globally hyperbolic spacetimes. That part deserves publication.

The Schwarzschild part does not hold up. The stress-test about Lemma 4.5 is correct. Take U*<0, V*=+infty and a regular point p=(U*,V_p,Omega_p) with Omega_p != Omega*. The set J^-(gamma_{U*,Omega*}) is not directed-sup-closed: choose a future-directed timelike chain e_n->p with U(e_n)Omega_p. Each e_n reaches gamma by rotating at constant r and then increasing V, so e_n in J^-(gamma). Global hyperbolicity makes p the supremum of {e_n}, and p not in J^-(gamma) because a causal curve with constant U has constant Omega. Hence the last equality in Lemma 4.5, \hat J^-(gamma)=J^-(gamma), is false. That breaks Proposition 4.2's null-infinity case, the inclusion T(p) subset T([U*,Omega*]) in Theorem 4.1, and the claimed order-reflection. In fact it looks like all [U*,Omega*] with the same U* have the same \hat J^-, so the angular parameter on I+ may collapse. This is not a minor gap; it changes the advertised model.

Minor issues: (4.16) is typographically garbled, and the proof of Theorem 4.1 outsources boundary mixed cases to 'analogous' checks. These are addressable, but secondary.

The paper is honest about its reliance on [10], and the self-citation is not a problem. The abstract overstates the result because the application is wrong as stated.

Who gets value: anyone working on causal boundaries or non-smooth Lorentzian geometry will want the Section 3 theorem. The Schwarzschild section needs substantial rework. I'd send it to a serious referee, with the stress-test attached. If Section 4 can be corrected, the paper could be good; as it stands the main advertised example fails.

Referee Report

2 major / 4 minor

Summary. The paper studies the directed completion of Lorentzian pre-length spaces and proves (Theorem 3.15) that, under an approximability, past-distinguishing, and a new 'supremum-compatibility' condition (SC), the future causal completion (C^+(X),⊆) is the directed completion of (X,≤) when causal pasts are closed. It then verifies (SC) for globally hyperbolic spacetimes (Propositions 3.16–3.17). The second half is devoted to the maximally extended Schwarzschild spacetime: the authors propose an explicit description of the directed completion (Theorem 4.1) in terms of radial null geodesics, with boundary pieces S^+, i^+, i^{+'}, I^+, I^{+'}, and prove that directed sets are represented by sets of the form \hat J^-(γ_{U^*,V^*,Ω^*}) (Proposition 4.2). The main technical work is a case-by-case study of directed sets according to the suprema of the Kruskal coordinates U and V.

Significance. If the results are correct, the paper gives a clean conceptual bridge between the order-theoretic directed completion and the classical Geroch–Kronheimer–Penrose future causal completion, with a concrete and nontrivial application to Schwarzschild. The Section 3 arguments are largely self-contained and rigorous; the reformulation of (SC) as the absence of TIPs admitting a least upper bound (Lemma 3.13) is a useful insight. The Schwarzschild analysis is ambitious and, modulo the gaps discussed below, would provide a new viewpoint on the causal boundary of Kruskal–Szekeres. The paper also usefully stresses that the general criterion of [10, Proposition 2.10] does not apply, since \hat D need not be directed.

major comments (2)

[§4.2.3, Lemma 4.5] The final assertion of Lemma 4.5 — that J^-(γ_{U^*,Ω^*}) is directed-sup-closed 'since V^*=+∞' — is not adequately proved and is load-bearing for Theorem 4.1. A directed subset E⊆J^-(γ_{U^*,Ω^*}) with a supremum p∈M_Krus need not admit a uniform V-bound on the points of γ used to connect elements of E to γ; the required V may tend to infinity as elements approach p. The proof must show that the angular estimate (4.19) (or a related argument) rules out the apparent possibility that p has U(p)=U^* and Ω(p)≠Ω^*. As written, the step is a gap, and the skeptic's proposed counterexample can be excluded only by an argument that is not supplied. Please expand this part with a complete proof of directed-sup-closedness.
[§4.1, proof of Theorem 4.1] The order-reflection and injectivity checks for pairs of boundary points from different pieces are relegated to 'the remaining mixed cases are analogous'. This is not a formality: the inclusion of boundary sets depends on the explicit forms of \hat J^-(γ) obtained in the various cases, and the comparison between, e.g., I^+ and S^+ or I^+ and i^+ is exactly where subtle mistakes could arise. Please provide a systematic verification for all mixed cases, or at least a representative case from each pair of boundary pieces, together with a clear statement of why the remaining cases follow by the same arguments.

minor comments (4)

[§4.2.3, Eq. (4.16)] The displayed estimate (4.16) contains corrupted LaTeX tokens ('/radicaltp/radicalvertex/radicalvertex√') and is unreadable. Please fix the formula and give a clean derivation of the inequality preceding (4.19).
[§4.2.3, after Eq. (4.15)] The notation 'for all p′∈D, p≤p′' after fixing p is slightly ambiguous: it should say that p is a lower bound for the tail {p′∈D : p≤p′}, and the subsequent estimates should explicitly state which quantities are evaluated at p′.
[§4.2.2, Lemma 4.4] The statement \hat D = {U≤0} would benefit from a short comment that this set is understood as a subset of M_Krus in the Kruskal coordinates, and that the equality is with the directed-sup-closure operation, not with the causal past of any point.
[§3.2, after Lemma 3.13] The paper relies heavily on [10] for the construction of the directed completion via transfinite recursion. Since this construction is not standard in all communities, a brief self-contained recap of the relevant theorem from [10] (or a precise statement of the first-step recursion) would improve readability.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the equivalence theorem and the Schwarzschild application are proved from stated hypotheses; self-citations supply definitions and standard facts only.

full rationale

The paper's central derivation is self-contained rather than circular. Theorem 3.15 proves directly that the future causal completion is the directed completion under the supremum-compatibility condition (SC): Lemma 3.12 establishes directed completeness of the collection of indecomposable past sets, condition (SC) is used explicitly to show that the embedding has the monotone convergence property, and the universal extension is constructed concretely via the formula in equation (3.4), with uniqueness following from the identity P = union over x in P of I^-(x). None of these steps assumes the conclusion. The condition (SC) is presented as a new hypothesis, not as a hidden restatement of the target result, and it is subsequently verified for smooth globally hyperbolic spacetimes through forward completeness in Propositions 3.16 and 3.17. The Schwarzschild characterization is likewise proved from the Kruskal metric by a case-by-case analysis: Proposition 4.2 reduces each directed set to the smallest lower directed-sup-closed set of a radial null geodesic using the estimates (4.13), (4.19), and (4.23), and Lemma 4.9 proves directed completeness of Y without importing the desired isomorphism. The citations to [10,15,23] provide the standard definition and construction of directed completion, not the equivalence theorem, while [1,16] are cited only for the standard fact that global hyperbolicity implies forward completeness; these self-citations are not load-bearing premises. The skeptical note about Lemma 4.5 concerns a possible mathematical error in an inclusion or estimate, which would be a correctness issue, not a circularity issue: no equality in the derivation chain is established by defining it into existence, and no fitted parameter is renamed as a prediction. Therefore the paper exhibits at most minor, non-load-bearing self-citation and no significant circularity.

Assumptions & free parameters 0 free parameters · 6 assumptions · 0 invented entities

The central result rests on standard Lorentzian geometry and order theory plus the specific hypotheses of the theorems. No parameters are fitted to data, and the boundary objects (i^+, I^+, S^+) are standard from the Penrose diagram literature, not newly postulated.

assumptions (6)

domain assumption Approximability (Definition 2.8)
Every point has a chronologically past/future point arbitrarily close; used in Lemmas 3.1, 3.5, 3.9 and Proposition 3.10.
domain assumption Past-distinguishing (Definition 2.7)
Needed to make (X,≤) a poset and to identify the chronological-inclusion order with ≤ via Lemma 3.9.
domain assumption Closed causal futures and pasts (Definition 2.9)
Used in Proposition 3.10 and Lemma 3.9 to pass from closures of chronological pasts to causal relations.
ad hoc to paper Supremum-compatibility condition (SC)
New assumption introduced in Section 3.2; required for the embedding ι to have the monotone convergence property, and shown to hold under forward completeness in Proposition 3.16.
domain assumption Forward completeness (Definition 2.10)
A Dedekind-type order completeness condition inherited from globally hyperbolic spacetimes; used to imply (SC).
domain assumption Global hyperbolicity of the Kruskal-Szekeres spacetime
The Kruskal-Szekeres spacetime is globally hyperbolic [8], so Theorem 3.15 applies; used throughout Section 4.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Ideal points, directed completion and the case of the maximally extended Schwarzschild spacetime." pith.science (2026). https://pith.science/paper/CJGBCMLB

@misc{pith2026260804718,
  author       = {Pith},
  title        = {Pith review of: Ideal points, directed completion and the case of the maximally extended Schwarzschild spacetime},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CJGBCMLB}},
  note         = {Machine review of arXiv:2608.04718}
}

read the original abstract

We study the directed completion of Lorentzian pre-length spaces, and we show that, under some natural assumptions which cover the case of smooth globally hyperbolic spacetimes, it coincides with the future causal completion of Geroch--Kronheimer--Penrose. Moreover, we provide applications of our findings by characterizing the directed completion of the Kruskal--Szekeres spacetime in terms of its radial null geodesics.

Figures

Figures reproduced from arXiv: 2608.04718 by the authors.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

23 extracted references · 20 canonical work pages

[10]

Gigli.Hyperbolic Banach spaces

N. Gigli.Hyperbolic Banach spaces. 2025. arXiv:2503.10467 [math.FA]. 25

arXiv 2025
[1]

Beran, M

T. Beran, M. Braun, M. Calisti, N. Gigli, R. J. McCann, A. Ohanyan, F. Rott, and C. S¨ amann.A nonlinear d’Alembert comparison theorem and causal differential calculus on metric measure spacetimes. 2024. arXiv:2408.15968 [math.DG]

arXiv 2024
[2]

Causal boundaries for general relativistic space–times

R. Budic and R. K. Sachs. “Causal boundaries for general relativistic space–times”. In: Journal of Mathematical Physics (N.Y.)15.8 (1974), pp. 1302–1309

work page 1974
[3]

Thec-completion of Lorentzian metric spaces

S. Burgos, J. L. Flores, and J. Herrera. “Thec-completion of Lorentzian metric spaces”. In:Classical Quantum Gravity40.20 (2023), Paper No. 205013, 25

work page 2023
[4]

Complete analytic extension of the symmetry axis of Kerr’s solution of Ein- stein’s equations

B. Carter. “Complete analytic extension of the symmetry axis of Kerr’s solution of Ein- stein’s equations”. In:Physical Review141.4 (1966), p. 1242

work page 1966
[5]

A Comparison of Whitehead’s and Einstein’s Formulæ

A. S. Eddington. “A Comparison of Whitehead’s and Einstein’s Formulæ”. In:Nature 113.2832 (1924), p. 192

work page 1924
[6]

Past-Future Asymmetry of the Gravitational Field of a Point Particle

D. Finkelstein. “Past-Future Asymmetry of the Gravitational Field of a Point Particle”. In:Physical Review110 (4 1958), pp. 965–967

work page 1958
[7]

Ideal points in space-time

R. Geroch, E. H. Kronheimer, and R. Penrose. “Ideal points in space-time”. In:Proceed- ings of the Royal Society. London. Series A327 (1972), pp. 545–567

work page 1972

Show all 23 references

[8]

Domain of dependence

R. Geroch. “Domain of dependence”. In:Journal of Mathematical Physics11 (1970), pp. 437–449

1970
[9]

Local Characterization of Singularities in General Relativity

R. Geroch. “Local Characterization of Singularities in General Relativity”. In:Journal of Mathematical Physics9.3 (1968), pp. 450–465

1968
[11]

Universality of the future chronological boundary

S. G. Harris. “Universality of the future chronological boundary”. In:Journal of Mathe- matical Physics39.10 (1998), pp. 5427–5445

1998
[12]

Causal completions as Lorentzian pre-length spaces

L. A. Hau, S. Burgos, and D. A. Solis. “Causal completions as Lorentzian pre-length spaces”. In:Gen. Relativity Gravitation54.9 (2022), Paper No. 108, 20

2022
[13]

Maximal Extension of Schwarzschild Metric

M. D. Kruskal. “Maximal Extension of Schwarzschild Metric”. In:Physical Review119 (5 1960), pp. 1743–1745

1960
[14]

Lorentzian length spaces

M. Kunzinger and C. S¨ amann. “Lorentzian length spaces”. In:Ann. Global Anal. Geom. 54.3 (2018), pp. 399–447

2018
[15]

Chain-complete posets and directed sets with applications

G. Markowsky. “Chain-complete posets and directed sets with applications”. In:Algebra Universalis6 (1976), pp. 53–68

1976
[16]

Ohanyan and M

A. Ohanyan and M. S´ alamo Candal.Timelike Ricci curvature lower bounds via optimal transport for Orlicz-type Lorentzian costs. 2026. arXiv:2604.22538

2026 arXiv
[17]

Asymptotic properties of fields and space-times

R. Penrose. “Asymptotic properties of fields and space-times”. In:Phys. Rev. Lett.10 (1963), pp. 66–68

1963
[18]

TheC 0-inextendibility of the Schwarzschild spacetime and the spacelike diameter in Lorentzian geometry

J. Sbierski. “TheC 0-inextendibility of the Schwarzschild spacetime and the spacelike diameter in Lorentzian geometry”. In:Journal of Differential Geometry108.2 (2018), pp. 319–378

2018
[19]

A construction of boundaries for field manifolds

B. G. Schmidt. “A construction of boundaries for field manifolds”. In:General Relativity and Gravitation1.3 (1971), pp. 269–280

1971
[20]

¨Uber das Gravitationsfeld eines Massenpunktes nach der Einstein- schen Theorie

K. Schwarzschild. “ ¨Uber das Gravitationsfeld eines Massenpunktes nach der Einstein- schen Theorie”. German. In:Sitzungsberichte der K¨ oniglich Preußischen Akademie der Wissenschaften1 (1916), pp. 189–196

1916
[21]

The abstract boundary—a new approach to singularities of manifolds

S. M. Scott and P. Szekeres. “The abstract boundary—a new approach to singularities of manifolds”. In:Journal of Geometry and Physics13.3 (1994), pp. 223–253

1994
[22]

On the singularities of a Riemannian manifold

G. Szekeres. “On the singularities of a Riemannian manifold”. In:Publicationes Mathe- maticae Debrecen7 (1960), pp. 285–301

1960
[23]

Dcpo-completion of posets

D. Zhao and T. Fan. “Dcpo-completion of posets”. In:Theoretical Computer Science 411.22-24 (2010), pp. 2167–2173. 26

2010

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [10]

Gigli.Hyperbolic Banach spaces

N. Gigli.Hyperbolic Banach spaces. 2025. arXiv:2503.10467 [math.FA]. 25

arXiv 2025

[2] [1]

Beran, M

T. Beran, M. Braun, M. Calisti, N. Gigli, R. J. McCann, A. Ohanyan, F. Rott, and C. S¨ amann.A nonlinear d’Alembert comparison theorem and causal differential calculus on metric measure spacetimes. 2024. arXiv:2408.15968 [math.DG]

arXiv 2024

[3] [2]

Causal boundaries for general relativistic space–times

R. Budic and R. K. Sachs. “Causal boundaries for general relativistic space–times”. In: Journal of Mathematical Physics (N.Y.)15.8 (1974), pp. 1302–1309

work page 1974

[4] [3]

Thec-completion of Lorentzian metric spaces

S. Burgos, J. L. Flores, and J. Herrera. “Thec-completion of Lorentzian metric spaces”. In:Classical Quantum Gravity40.20 (2023), Paper No. 205013, 25

work page 2023

[5] [4]

Complete analytic extension of the symmetry axis of Kerr’s solution of Ein- stein’s equations

B. Carter. “Complete analytic extension of the symmetry axis of Kerr’s solution of Ein- stein’s equations”. In:Physical Review141.4 (1966), p. 1242

work page 1966

[6] [5]

A Comparison of Whitehead’s and Einstein’s Formulæ

A. S. Eddington. “A Comparison of Whitehead’s and Einstein’s Formulæ”. In:Nature 113.2832 (1924), p. 192

work page 1924

[7] [6]

Past-Future Asymmetry of the Gravitational Field of a Point Particle

D. Finkelstein. “Past-Future Asymmetry of the Gravitational Field of a Point Particle”. In:Physical Review110 (4 1958), pp. 965–967

work page 1958

[8] [7]

Ideal points in space-time

R. Geroch, E. H. Kronheimer, and R. Penrose. “Ideal points in space-time”. In:Proceed- ings of the Royal Society. London. Series A327 (1972), pp. 545–567

work page 1972