REVIEW 1 major objections 5 minor 1 cited by
The metric geometry of singularity types
T0 review · 1 major / 5 minor · reviewed 2026-08-14 · deepseek-v4-flash
Pith's one-line read The paper introduces a pseudometric on singularity types that becomes a complete metric on positive-mass subspaces, with atoms equal to relative full mass classes.
desk verdict A genuinely useful metric on singularity types, with solid positive-mass theorems, but the abstract overstates the zero-mass atom characterization by relying on the still-open equality P=C. 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 engine is the map $r[\,\cdot\,]:\mathcal S(X,\theta)\to\mathcal R(X,\theta)$ that attaches to each singularity type the geodesic ray with minimal singularity type, along with the ceiling operator $C$ whose fixed points are the model potentials. The metric is the chordal limit $d_{\mathcal S}([\psi],[\chi])=\lim_{t\to\infty} d_1(r[\psi]_t,r[\chi]_t)/t$, and the paper repeatedly uses the identity $P[\psi]=C(\psi)$ for positive mass, so model potentials are exactly envelope-fixed points. This machinery turns singularity-type questions into geodesic-ray questions, where completeness is already available.
What would settle it
Construct a $d_{\mathcal S}$-Cauchy sequence $[u_j]$ in $\mathcal S_\delta(X,\theta)$ whose decreasing envelope limit $v=\lim_j C(v_j)$ is not a fixed point of $C$ or has total mass below $\delta$; Theorem 1.1 predicts that no such sequence can exist. On the elliptic ruled surface example showing incompleteness for $\delta=0$, the analogous sequence loses mass precisely at the limit, so the comparison is explicit.
Extended reading notes
Core claim
The paper establishes that $\mathcal S(X,\theta)$ admits a pseudometric $d_{\mathcal S}$ whose zero classes are precisely the relative full mass classes: $d_{\mathcal S}([\psi],[\chi])=0$ iff $C(\psi)=C(\chi)$, equivalently the associated geodesic rays coincide. On $\mathcal S_\delta(X,\theta)$, the subspace of types with $\int_X\theta_u^n\ge\delta>0$, this pseudometric is complete. The paper further proves a volume diamond inequality $\int_X\theta_u^n+\int_X\theta_v^n\le \int_X\theta_{\max(u,v)}^n+\int_X\theta_{P(u,v)}^n$ when $P(u,v)\in\mathrm{PSH}(X,\theta)$, and derives from it two applications: $d_{\mathcal S}$-convergence forces eventual inclusion of multiplier ideal sheaves, and solutions of complex Monge-Ampère equations with prescribed singularity converge in capacity when the prescribed types converge.
Load-bearing premise
The completeness proof assumes an imported identity: for every positive-mass potential, the envelope $P[\psi]$ equals the ceiling $C(\psi)$; if that external theorem were false, the Cauchy-sequence limit argument in Theorem 4.9 would not go through.
Editorial extensions
If this is right
- Increasing sequences of $\theta$-psh potentials satisfy $d_{\mathcal S}([u_j],[u])\to0$, and $d_{\mathcal S}$-convergence can be characterized by sandwiching a sequence between increasing and decreasing approximating sequences.
- For any $\delta>0$, every $d_{\mathcal S}$-Cauchy sequence in $\mathcal S_\delta(X,\theta)$ converges to a singularity type whose total Monge-Ampère mass is at least $\delta$.
- The volume diamond inequality $\int_X\theta_u^n+\int_X\theta_v^n\le\int_X\theta_{\max(u,v)}^n+\int_X\theta_{P(u,v)}^n$ holds whenever the rooftop $P(u,v)$ is $\theta$-psh, and it is an identity in complex dimension one.
- If $d_{\mathcal S}([u_j],[u])\to0$, then the multiplier ideal sheaf inclusion $\mathcal J[u]\subseteq\mathcal J[u_j]$ holds for all sufficiently large $j$.
- Prescribed singularity types that converge in $d_{\mathcal S}$ make the solutions of the corresponding complex Monge-Ampère equations converge in capacity and in $L^1$.
Reading between the lines
- Editorial inference: the $d_{\mathcal S}$-topology gives a quantitative notion of approximation for singularity types, so the effectiveness of kernel-type or envelope-type approximation schemes could be measured by rates of $d_{\mathcal S}$-convergence rather than only by stabilization of cohomological invariants.
- Editorial inference: if the zero-mass identity $P[\psi]=C(\psi)$ stated as Conjecture 2.5 holds, the completeness and diamond arguments may extend below the positive-mass threshold, potentially making selected subspaces of $\mathcal S(X,\theta)$ complete even when the full space is not.
- Editorial inference: a local analog of $d_{\mathcal S}$ on singularity germs of plurisubharmonic functions would turn the multiplier-ideal semicontinuity theorem into a quantitative strong-openness statement, though the paper only records this as a motivating question.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper introduces a pseudometric d_S on the space S(X,θ) of singularity types of θ-psh potentials on a compact Kähler manifold with a big cohomology class. The construction embeds S(X,θ) into the chordal metric space of finite-energy geodesic rays and pulls back the metric d_1^c. The main results are: (i) d_S is a pseudometric whose zero-distance relation is characterized by equality of the ceiling operator C (Theorem 3.3); (ii) for δ>0 the subspace S_δ(X,θ) is complete (Theorem 1.1/4.9), while the full space is incomplete (Section 4.2); (iii) a volume diamond inequality for non-pluripolar masses (Theorem 1.2/5.4); and applications to semicontinuity of multiplier ideal sheaves (Theorem 1.3/6.1) and to stability of complex Monge–Ampère equations with prescribed singularity type (Theorem 1.4/7.1). The proofs are detailed and rely on prior work of the authors, especially the identity P[φ]=C(φ) for positive-mass potentials.
Significance. If the main results hold, this paper provides a natural metric topology on singularity types, filling a gap in the literature where convergence of singularity types was previously treated only in ad-hoc ways. The completeness theorem for positive-mass subspaces and the stability applications are substantial and will likely be useful in pluripotential theory and transcendental algebraic geometry. The paper is carefully written and contains numerous proved lemmas rather than black-box statements. However, the advertised atom characterization in the abstract is stronger than what is formally established, as detailed in the major comment; this does not undermine the positive-mass results but requires a correction of the claimed scope.
major comments (1)
- [Abstract and §1; Theorem 3.3; Conjecture 2.5] The headline assertion that the atoms of d_S are 'exactly the relative full mass classes' (Abstract and §1, p. 2) is stronger than what is proved. Theorem 3.3 shows d_S([ψ],[χ])=0 iff C(ψ)=C(χ), and the identification of C-equivalence classes with relative full mass classes requires the identity P[φ]=C(φ). For positive mass this is imported from [DDL2, Remark 2.5, Theorem 3.12] and is used in Lemma 4.3, Proposition 4.6, Corollary 4.7, Proposition 4.8 and Theorem 4.9. For zero mass the identity is not proved and is explicitly left open as Conjecture 2.5; zero-mass singularity types do occur in S(X,θ) (see φ_2 in §4.2). Consequently the advertised atom characterization is not established in full generality. Please either prove the zero-mass case or reformulate the abstract/introduction to state that for positive mass the atoms are the relative full mass classes, while in general the atoms are exactly the fibers of C, with the identification to relative full mass classes being conjectural (Conjecture 2.5). This change is needed for the central claim to match the formal results.
minor comments (5)
- [§1, p. 2] In the introduction, the definition of the envelope reads 'P [φ] := sup {v ∈ PSH(X,θ ) : [ v] ≤ [u],v ≤ 0}', using both φ and u; this is a mismatch and should be corrected (likely [v] ≤ [φ]).
- [§4.2] In the computation of the mixed mass of φ_t, the inequality '≥ {θ}.{η} = ∫_C η >0' would be clearer if the authors explicitly noted that {θ}.{η} is an intersection number and that ∫_C η >0 because η is positive along the curve C.
- [§2.1, Lemma 2.4] In the proof of Lemma 2.4, the sentence 'we used [DDL2, Proposition 2.1, Theorem 2.2]' compresses two different external facts into one citation; a more detailed pointer would improve readability.
- [§4, Proposition 4.2] The geometric series bound in inequality (18), '∑_{k≥j} 1/C^{k+j−1} ≤ 1/C^{j−1} C/(C−1)', is correct but the intermediate step ∑_{k≥j} C^{-(k+j-1)} = C^{-(2j-1)}/(1-1/C) could be displayed to help the reader verify the estimate.
- [§5, Theorem 5.6] The notation w_j := P(u_j,u_{j+1},...) in (30) denotes an infinite envelope; the accompanying text explains it as a decreasing limit of finite envelopes, but stating this explicitly in the theorem statement would prevent confusion.
Circularity Check
No circular derivation: d_S is genuinely constructed and its zero-distance classes are proved to equal C-equivalence; the only caveat is a zero-mass atom claim left to Conjecture 2.5, which is an unsupported assertion, not a circular reduction.
full rationale
dS is not defined to have the claimed atoms; it is pulled back from the chordal metric on geodesic rays (Section 2.2, (8)), and the equivalence dS([ψ],[χ])=0 iff C(ψ)=C(χ) is proved in Proposition 3.2 and Theorem 3.3 by showing r[ψ]∞=C(ψ) and using the metric completeness of the ray space. The ceilings C(ψ) are defined independently by a mixed-mass envelope family (4), not by dS. The completeness of Sδ(X,θ) (Theorem 4.9) is a proof from Lemma 4.3, Propositions 4.6 and 4.8, and the reduction dS([vj],[C(vj)])=0 follows from Theorem 3.3. The most delicate input is the positive-mass identity P[φ]=C(φ), imported from [DDL2, Remark 2.5, Theorem 3.12] and used in Lemmas 2.4 and 4.3; this is a load-bearing self-citation, but it is a published theorem with the needed hypotheses, not a parameter fitted in this paper, and the paper explicitly separates the unproved zero-mass case as Conjecture 2.5. Consequently the abstract's statement that atoms are exactly the relative full mass classes is not fully supported in the zero-mass case—if P[φ]=C(φ) fails there, the atom description based on C-equivalence would not match the P-based relative full mass classes. This is a correctness/incompleteness caveat, not circularity: no equation in the paper is equivalent to an input by construction, and no fitted quantity is renamed as a prediction.
Assumptions & free parameters
assumptions (6)
- standard math Non-pluripolar Monge-Ampère products are well-defined for θ-psh potentials in big cohomology classes and satisfy mass monotonicity ([BEGZ10], [WN19]).
- standard math The space (E^1(X,θ), d_1) is a complete geodesic metric space and finite-energy geodesics exist ([DDL3, Theorem 1.1]).
- standard math For positive mass potentials φ, the envelope P[φ] equals the ceiling C(φ) ([DDL2, Remark 2.5, Theorem 3.12]).
- standard math Guan-Zhou strong openness: for increasing sequences of psh functions u_j ↑ u, the multiplier ideal sheaves stabilize ([GZh15], [GZh16]).
- standard math Domination principle and relative full mass characterization via envelopes ([DDL1, Proposition 2.4], [DDL2, Sections 2-3]).
- standard math Skoda's uniform integrability theorem for negative psh functions with zero supremum ([Zer01], [GZ17, Theorem 2.50]).
Cite this review
Pith. "Pith review of The metric geometry of singularity types." pith.science (2026). https://pith.science/paper/Y3EBVUTB
@misc{pith2026190900839,
author = {Pith},
title = {Pith review of: The metric geometry of singularity types},
year = {2026},
howpublished = {\url{https://pith.science/paper/Y3EBVUTB}},
note = {Machine review of arXiv:1909.00839}
}
abstract
Let $X$ be a compact K\"ahler manifold. Given a big cohomology class $\{\theta\}$, there is a natural equivalence relation on the space of $\theta$-psh functions giving rise to $\mathcal S(X,\theta)$, the space of singularity types of potentials. We introduce a natural pseudometric $d_{\mathcal {S}}$ on $\mathcal S(X,\theta)$ that is non-degenerate on the space of model singularity types and whose atoms are exactly the relative full mass classes. In the presence of positive mass we show that this metric space is complete. As applications, we show that solutions to a family of complex Monge-Amp\`ere equations with varying singularity type converge as governed by the $d_\mathcal S$-topology, and we obtain a semicontinuity result for multiplier ideal sheaves associated to singularity types, extending the scope of previous results from the local context.
Forward citations
Cited by 1 Pith paper
-
Complex Hessian equations with prescribed singularity on compact K\"ahler manifolds
The total Hessian mass is monotone in singularity type, and Hessian equations H_m(u)=µ have unique solutions in prescribed singularity classes on compact Kähler manifolds.
Reference graph
Works this paper leans on
-
[1]
E. Bedford, B.A. Taylor, The Dirichlet problem for a complex Monge-Amp\`ere equation, Invent. Math. 37 (1976), no. 1, 1--44
work page 1976
-
[2]
E. Bedford, B.A. Taylor, A new capacity for plurisubharmonic functions , Acta Math. 149 (1982), 1--40
work page 1982
-
[3]
E. Bedford, B.A. Taylor, Fine topology, Silov boundary, and (i )^n , J. Funct. Anal. 72 (1987), no. 2, 225--251
work page 1987
-
[4]
R.J. Berman, From Monge-Amp\`ere equations to envelopes and geodesic rays in the zero temperature limit, Math. Z., 2019, 291, 1--2, 365--394
work page 2019
- [5]
- [6]
-
[7]
B. Berndtsson, The openness conjecture and complex Brunn-Minkowski inequalities, Complex geometry and dynamics, 29--44, Abel Symp., 10, Springer, Cham, 2015
work page 2015
-
[8]
Z. Blocki, The complex Monge-Amp\`ere operator in pluripotential theory, lecture notes, http://gamma.im.uj.edu.pl/ blocki/publ/ln/wykl.pdf http://gamma.im.uj.edu.pl/\ blocki/publ/ln/wykl.pdf
Show all 36 references
-
[9]
Blocki, S
Z. Blocki, S. Ko odziej, On regularization of plurisubharmonic functions on manifolds, Proc. Amer. Math. Soc., 2007, Vol 135, 7, 2089--2093
2007
-
[10]
Boucksom, On the volume of a line bundle
S. Boucksom, On the volume of a line bundle. Internat. J. Math. 13 (2002), no. 10, 1043--1063
2002
-
[11]
Boucksom, Divisorial Zariski decompositions on compact complex manifolds, Ann
S. Boucksom, Divisorial Zariski decompositions on compact complex manifolds, Ann. Sci. \'Ecole Norm. Sup. (4) 37 (2004), no. 1, 45--76
2004
-
[12]
Boucksom, P
S. Boucksom, P. Eyssidieux, V. Guedj, A. Zeriahi, Monge-Amp\`ere equations in big cohomology classes, Acta Math. (2010), Volume 205, Issue 2, pp 199--262
2010
-
[13]
Cao, Numerical dimension and a Kawamata-Viehweg-Nadel-type vanishing theorem on compact K\"ahler manifolds
J. Cao, Numerical dimension and a Kawamata-Viehweg-Nadel-type vanishing theorem on compact K\"ahler manifolds. Compos. Math. 150 (2014), no. 11, 1869--1902
2014
-
[14]
Darvas, Weak geodesic rays in the space of K\"ahler potentials and the class E(X, ) , J
T. Darvas, Weak geodesic rays in the space of K\"ahler potentials and the class E(X, ) , J. Inst. Math. Jussieu 16 (2017), no. 4, 837--858
2017
-
[15]
Darvas, The Mabuchi geometry of finite energy classes, Adv
T. Darvas, The Mabuchi geometry of finite energy classes, Adv. Math. 285 (2015), 182--219
2015
-
[16]
Darvas, E
T. Darvas, E. Di Nezza, C.H. Lu, On the singularity type of full mass currents in big cohomology classes, Compos. Math. 154 (2018), no. 2, 380--409
2018
-
[17]
Darvas, E
T. Darvas, E. Di Nezza, C.H. Lu, Monotonicity of non-pluripolar products and complex Monge-Ampere equations with prescribed singularity, Analysis & PDE, Vol. 11 (2018), No. 8, 2049--2087
2018
-
[18]
Darvas, E
T. Darvas, E. Di Nezza, C.H. Lu, L^1 metric geometry of big cohomology classes, Annales de L'Institut Fourier, 68, no 7 (2018), p. 3053--3086
2018
-
[19]
Darvas, E
T. Darvas, E. Di Nezza, C.H. Lu, Log-concavity of volume and complex Monge-Amp\`ere equations with prescribed singularity, arXiv:1807.00276
-
[20]
Darvas, C.H
T. Darvas, C.H. Lu, Geodesic stability, the space of rays, and uniform convexity in Mabuchi geometry, arXiv:1810.04661
-
[21]
Demailly, J
J.-P. Demailly, J. Kollar, Semi-continuity of complex singularity exponents and K\"ahler-Einstein metrics on Fano orbifolds, Ann. Sci. \'Ecole Norm . Sup. 34 (2001), 525--556
2001
-
[22]
Demailly, Regularization of closed positive currents and intersection theory, J
J.-P. Demailly, Regularization of closed positive currents and intersection theory, J. Algebraic Geom. 1 (1992) 361--409
1992
-
[23]
Demailly, On the cohomology of pseudoeffective line bundles
J.-P. Demailly, On the cohomology of pseudoeffective line bundles. Complex geometry and dynamics, 51--99, Abel Symp., 10, Springer, Cham, 2015
2015
-
[24]
ahler cone of a compact K\
J.-P. Demailly, M. Paun, Numerical characterization of the K\"ahler cone of a compact K\"ahler manifold. Ann. of Math. (2) 159 (2004), no. 3, 1247--1274
2004
-
[25]
Demailly, T
J.-P. Demailly, T. Peternell, M. Schneider, Compact complex manifolds with numerically effective tangent bundles, J. Algebraic Geom., 3 (1994), 295--345
1994
-
[26]
Q. Guan, X. Zhou, A proof of Demailly's strong openness conjecture, Ann. of Math. (2) 182 (2015), no. 2, 605--616
2015
-
[27]
Q. Guan, X. Zhou, Effectiveness of Demailly's strong openness conjecture and related problems. Invent. Math. 202 (2015), no. 2, 635--676
2015
-
[28]
Guedj, A
V. Guedj, A. Zeriahi, Degenerate Complex Monge-Amp\`ere Equations, EMS Tracts in Mathematics, vol. 26, (2017)
2017
-
[29]
Guedj, C.H
V. Guedj, C.H. Lu, A. Zeriahi, Plurisubharmonic envelopes and supersolutions , arXiv:1703.05254
-
[30]
Hartshorne, Algebraic Geometry, Springer Science and Business Media, (1977)
R. Hartshorne, Algebraic Geometry, Springer Science and Business Media, (1977)
1977
-
[31]
Hiep, The weighted log canonical threshold
P.H. Hiep, The weighted log canonical threshold. C. R. Math. Acad. Sci. Paris 352 (2014), no. 4, 283--288
2014
-
[32]
Lempert, Modules of square integrable holomorphic germs, Analysis meets geometry, 311--333, Trends Math., Birkhäuser/Springer, Cham, 2017
L. Lempert, Modules of square integrable holomorphic germs, Analysis meets geometry, 311--333, Trends Math., Birkhäuser/Springer, Cham, 2017
2017
-
[33]
J. Ross, D. Witt Nystr\"om, Analytic test configurations and geodesic rays, Journal of Symplectic Geometry Volume 12, Number 1 (2014), 125--169
2014
-
[34]
Xia, Mabuchi geometry of big cohomology classes with prescribed singularities, arXiv:1907.07234
M. Xia, Mabuchi geometry of big cohomology classes with prescribed singularities, arXiv:1907.07234
1907 arXiv
-
[35]
Witt-Nystr\"om, Monotonicity of non-pluripolar Monge-Amp\`ere measures, Indiana University Mathematics Journal, 2019, 68, 2, 579--591
D. Witt-Nystr\"om, Monotonicity of non-pluripolar Monge-Amp\`ere measures, Indiana University Mathematics Journal, 2019, 68, 2, 579--591
2019
-
[36]
Zeriahi, Volume and capacity of sublevel sets of a Lelong class of psh functions
A. Zeriahi, Volume and capacity of sublevel sets of a Lelong class of psh functions. Indiana Univ. Math. J. 50 (2001), no. 1, 671--703. footnotesize
2001
Reviewed August 14, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.