REVIEW 2 major objections 4 minor 1 cited by
Uniqueness of tangent currents for positive closed currents
T0 review · 2 major / 4 minor · reviewed 2026-08-08 · deepseek-v4-flash
Pith's one-line read Fast tube convergence pins a unique tangent current to a positive closed current along a submanifold.
desk verdict New uniqueness criterion for tangent currents that is worth refereeing; Section 6 has a repairable gap at the zero section and Theorem 1.13 depends on unpublished work. 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 proof is carried by the generalized Lelong numbers ν_j(T,B,r,τ,h), defined as normalized masses of τ_*T on tubes over B, together with strongly admissible maps τ that straighten a neighbourhood of B into the normal bundle E and are identity on V to first order. The main identity is the Lelong–Jensen formula for vector bundles (Theorem 2.8), which expresses the difference ν_j(r)-ν_j(r/2) as a vertical boundary term plus a local mass κ_j on the corona tube. A technique of admissible estimates compares κ_j with positive global mass indicators K_{j,q}; a key global inequality bounds these indicators, and two classical lemmas on functions on a punctured disk convert the resulting integrability bounds into convergence of the coefficients of the rescaled currents as λ→∞. The finite-cover version runs the same argument on each domain of a cover and matches the limits on overlaps.
What would settle it
To disprove the conditional claim, one would need a positive closed current in the stated approximability class for which both integrability conditions hold, yet two different tangent currents arise from two different strongly admissible maps. A concrete starting point is a product construction whose fibre factor is a direction-dependent tangent cone at a point, chosen so that its average-mean oscillation decays faster than the integrability thresholds while the tangent cone remains non-unique; comparing the limits for two maps would settle whether the theorem's hypotheses are sufficient.
Extended reading notes
Core claim
The central claim, stated as Theorem 1.9, is that uniqueness of tangent currents follows from an integrability condition on the generalized Lelong numbers. For a positive closed (p,p)-current T on a complex manifold X, a Kähler submanifold V of dimension l, and a relatively compact piecewise $C^{2}$ domain B⊂V, assume T=T^+-T^- near B with T^± in the class CL_{p;1,1}(B), meaning each is weakly approximable by smooth closed forms with uniformly bounded $C^{1}$ norms near the boundary of B. If for every j between max(0,l-p) and min(l,k-p) either the square-root integral of the difference between ν_j at radii r and r/2 or the plain integral of the difference between ν_j(r) and the limit ν_j is finite, then T admits a unique tangent current along B. The paper also establishes a finite-cover version with conditions on local mass indicators and proves that for the current of integration over a pure-codimension-p analytic set S the average means converge with rate O(r^ρ), so uniqueness holds there and the convergence speed is quantified.
Load-bearing premise
The proof rests on the assumption that, near the boundary of B, the current can be written as a difference of two positive closed currents that are limits of smooth forms with controlled size and controlled first derivatives; without that, the key formulas and mass bounds do not apply.
Editorial extensions
If this is right
- For any positive closed current in the stated class satisfying either integrability condition, the rescaled family converges as λ→∞ rather than merely along subsequences, so the tangent current is a uniquely defined invariant of T along B.
- The finite-cover theorem makes the criterion checkable locally: one only needs estimates on tube masses over pieces of a cover of B, which is useful when the singular support of T meets the boundary of B.
- Every current of integration over a complex analytic set S whose intersection with B is relatively compact in B admits a unique tangent current along B, with the explicit rate O(r^ρ) for the approach of average means to their limits.
- Because the generalized Lelong numbers are independent of the strongly admissible map, the uniqueness criterion is intrinsic even though the proof tracks approximations that are not holomorphic.
- The result extends the classical point-based uniqueness criterion to positive-dimensional base sets, so it applies to excess-intersection situations where the expected intersection dimension is larger than the classical one.
Reading between the lines
- The integrability conditions are sufficient, not necessary; a natural test is whether much slower, Dini-type decay of the tube averages still forces uniqueness, since the two key lemmas only need specific weighted integrability.
- The equivalence between the κ_j conditions and the global mass indicators suggests a practical certificate: if one can bound κ_j by a small power of r on each piece of a cover, uniqueness follows without computing the generalized Lelong numbers.
- For analytic sets, the polynomial rate O(r^ρ) should imply stability of the tangent current under holomorphic deformations of S, a consequence not stated in the paper.
- The remark about weakening the Kähler assumption indicates that the true ingredients are the strongly admissible map and boundary C^1 control; testing the theorem on non-Kähler examples would clarify whether the integrability condition alone is the operative hypothesis.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper claims a uniqueness criterion for tangent currents of positive closed currents along a Kähler submanifold. Theorem 1.9 states that if T = T^+ - T^- with T^± in the class CL_{p;1,1}(B), and if the averaged generalized Lelong numbers ν_j(T,B,r) converge to ν_j(T,B) fast enough so that the differences are integrable against r^{-1} dr, then the family (A_λ)_* τ_* T has a unique limit as |λ| → ∞. Theorem 1.10 gives a local version over a finite cover of B by piecewise C^2 domains. Theorem 1.13 shows that for currents of integration over complex analytic sets the differences ν_j(r) - ν_j have order O(r^ρ), so the criterion applies. The proofs reduce the integrability conditions to growth estimates of global mass indicators using Lelong–Jensen formulas, pass to local coefficient estimates, and then use the Blel–Demailly–Mouzali limiting argument for functions on the punctured disc.
Significance. If the main result is correct, it is a natural and valuable generalization of the Blel–Demailly–Mouzali criterion from points to higher-dimensional bases, with concrete applications to intersection theory. The paper does not fit free parameters and the analytic-set rate in Theorem 1.13 is a useful quantitative contribution. The arguments are not self-contained and rely heavily on the first author's long preprint [20], but the cited statements are precise enough to be checked. The main obstruction to accepting the paper as written is a gap in the final convergence argument near the zero section, which is exactly the region where tangent currents may carry mass.
major comments (2)
- [Section 6, End of the proof of Theorem 1.9] The proof only establishes weak convergence of T_λ on E|B ∖ B, not on a neighborhood of B. The text explicitly says 'We only need to prove that Tλ converge weakly locally in E|B∖B' and then takes test functions supported in a relatively compact open set in E|B∖B. This is not sufficient: tangent currents are defined on all of π⁻¹(B) and may be supported on B. For example, if T = [V], then Remark 1.6 and Theorem 1.13 imply that the hypotheses of Theorem 1.9 hold, but T_λ = [V] for every λ; this current is zero on E|B∖B and equals [V] on B. Convergence away from B cannot distinguish [V] from 0. The same gap affects the proofs of Theorem 1.10 and the final statement of Theorem 1.13. The authors need either to prove weak convergence for test forms supported near B or to prove an extension principle, for instance that any two V-conic subsequential limits that agree on E|B∖B and have the same compactly supported cohomology class must coincide; neither argument is present.
- [Proposition 4.7, proof of assertion (a)] After equation (4.10), the proof bounds the sum of K_{j,k-p-j}(T,B,r/2,r) by a constant times the sum of |κ_j(T,B,r/2,r)| plus c r^{1/4}, citing what appears to be Proposition 4.5 but should be Proposition 4.6(1). However, Proposition 4.6 is stated for T in the positive class CL_{p;1,1}(U,W), while the current T in Theorem 1.9 is a signed difference T^+ - T^- with T^± in that class. The displayed reduction therefore needs an explicit linearity argument: apply the estimates to T^+ and T^- separately, then use the Lelong–Jensen formula for each positive part to control κ_j(T^±) by differences of the corresponding ν_j(T^±,r). As written, the step is not justified and is load-bearing for the reduction of Theorem 1.9 to the mass-indicator integrability.
minor comments (4)
- [Section 7, equation (7.3)] Equation (7.3) appears to be incomplete: it reads 'CpSq ∩ BverTubepB, 1q' with no asserted relation or equality. From the surrounding argument, the intended statement is presumably that this intersection is empty for r small enough, since this is what is used later to remove the first term in the formula for Mver(r1,r2). Please correct this.
- [Proposition 4.7(a)] The reference to 'Proposition 4.5' in the proof of assertion (a) seems to be a typo; the inequality relating K_{j,k-p-j}(T,B,r/2,r) and |κ_j(T,B,r/2,r)| is Proposition 4.6(1).
- [Throughout] There are minor grammatical and typographical issues, for example 'Let T be a positive closed currents' in the abstract, 'for for all' near (1.10), and 'τ is a admissible map' in Theorem 1.13. These do not affect the mathematics.
- [References] Theorem 1.13 relies on the construction of cylindrical cones from the item marked 'In progress' in the references. Since this is a load-bearing external tool for that theorem, the authors should either provide a stable reference or include a self-contained statement of the needed construction.
Circularity Check
No significant circularity: the integrability hypotheses are not defined in terms of the uniqueness conclusion, and the heavy reliance on prior work is support-based rather than definitional.
full rationale
The paper's main implication (Theorems 1.9 and 1.10) is a genuine derivation: conditions (a) and (b) are integrability conditions on differences of generalized Lelong numbers, and the conclusion is that the dilated currents converge. The proof reduces these conditions to decay of mass indicators, then applies the Blel-Demailly-Mouzali limiting lemma; no fitted parameter is renamed as a prediction, and no conclusion is built into an assumption. The paper does lean substantially on the first author's prior work [20] for the CL_{p;1,1} classes, the Lelong-Jensen formula (Theorem 2.8), the mass estimate (Theorem 4.2), and the admissible-estimates technique. This is heavy self-citation, but the cited results are prior mathematical theorems with stated assumptions that do not include the uniqueness conclusion, so the new implication is not forced by the citation alone. Theorem 1.13 uses the cylindrical cone construction from [22], which is listed as unpublished and by the same first author; this is a legitimate concern about support and completeness, but not circularity, because the construction is not assumed to prove the theorem. The skeptic's objection that Section 6 proves convergence only on E|B \ B and does not directly control tangent currents supported on the zero section is a possible gap in the proof of uniqueness, not a circular step: it concerns whether the argument establishes the stated conclusion, not whether the conclusion is equivalent to the input by construction. Overall, no circular step can be exhibited from the paper's own equations.
Assumptions & free parameters
assumptions (7)
- domain assumption Lelong-Jensen formula for approximable currents (Theorem 2.8) holds for T in the class CL_{p;1,1}(B).
- domain assumption Uniform mass bound K_{j,q}(T,r) < c for currents in \tilde{CL}_{p;1,1}(U,W) (Theorem 4.2).
- domain assumption Existence of strongly admissible maps along B for X Kähler (Proposition 2.7) and along each domain B_i.
- domain assumption The class CL_{p;1,1}(B) contains integration currents over analytic sets (Remark 1.6).
- domain assumption Theorem 1.7 (existence of generalized Lelong numbers and tangent currents for approximable T) holds.
- standard math Stokes' theorem and Sard's theorem apply to boundaries of tubes and analytic sets.
- standard math The classical BDM limiting lemmas (6.2, 6.3) and volume estimate (7.1) are valid.
Cite this review
Pith. "Pith review of Uniqueness of tangent currents for positive closed currents." pith.science (2026). https://pith.science/paper/KXZKDHIU
@misc{pith2026250206532,
author = {Pith},
title = {Pith review of: Uniqueness of tangent currents for positive closed currents},
year = {2026},
howpublished = {\url{https://pith.science/paper/KXZKDHIU}},
note = {Machine review of arXiv:2502.06532}
}
abstract
Let $X$ be a complex manifold $X$ of dimension $k,$ and let $V\subset X$ be a K\"ahler submanifold of dimension $l,$ and let $B\subset V$ be a piecewise $\mathcal{C}^2$-smooth domain. Let $T$ be a positive closed currents of bidegree $(p,p)$ in $X$ such that $T$ satisfies a mild reasonable assumption in a neighborhood of $\partial B$ in $X$ and that the $j$-th average mean $\nu_j(T,B,r)$ for every $j$ with $\max(0,l-p)\leq j\leq\min(l,k-p)$ converges sufficiently fast to the $j$-th generalized Lelong number $\nu_j(T,B)$ as $r$ tends to $0$ so that $r^{-1}(\nu_j(T, B,r)-\nu_j( T,B))$ is locally integrable near $r=0.$ Then we show that $T$ admits a unique tangent current along $B.$ A local version where we replace the condition of $T$ near $B$ by the conditions on a finite cover of $B$ by piecewise $\mathcal{C}^2$-smooth domains in $V$ is also given. When $T$ is a current of integration over a complex analytic set, we show that $\nu_j(T,B,r)-\nu_j(T,B)=O(r^\rho)$ for some $\rho>0,$ and hence this condition is satisfied. Our result may be viewed as a natural generalization of Blel-Demailly-Mouzali's criterion from the case $l=0$ to the case $l>0.$ The result has applications in the intersection theory of positive closed currents.
Forward citations
Cited by 1 Pith paper
-
Continuous local potential functionals and the Dinh-Sibony product
On any complex manifold, the Dinh-Sibony product of three positive closed currents is well defined and associative when the first current has continuous local potential functionals and the other two satisfy Condition (I).
Reference graph
Works this paper leans on
-
[22]
Nguyˆ en, Viˆ et-Anh: Thie type theorems and the general ized Lelong numbers for complex analytic sets. In progress
-
[20]
Nguyˆ en, Viˆ et-Anh: Positive plurisubharmonic curre nts: Generalized Lelong numbers and Tangent theorems. arXiv:2111.11024v2, 241 pages
-
[1]
Alessandrini, Lucia; Bassanelli, Giovanni: Lelong num bers of positive plurisubharmonic currents. Results Math. 30 (1996), no. 3-4, 191–224
work page 1996
-
[2]
(French) [Tangent cone of a nonnegative closed p1, 1q current] C
Blel, Mongi: Cˆ one tangent ` a un courant positif ferm´ e d e type p1, 1q. (French) [Tangent cone of a nonnegative closed p1, 1q current] C. R. Acad. Sci. Paris S´ er . I Math.309 (1989), no. 8, 543-546
work page 1989
-
[3]
(French) [Existence of the tangent cone of a closed positive current] Ark
Blel, Mongi; Demailly , Jean-Pierre; Mouzali, Mokhtar: Sur l’existence du cˆ one tangent ` a un courant positif ferm´ e. (French) [Existence of the tangent cone of a closed positive current] Ark. Mat. 28 (1990), no. 2, 231–248
work page 1990
-
[4]
Manuscript available at http://www-fourier.ujf-grenoble.fr/„demailly/books.html, (2012)
Demailly , Jean-Pierre: Complex Analytic and Different ial Geometry . Manuscript available at http://www-fourier.ujf-grenoble.fr/„demailly/books.html, (2012)
work page 2012
-
[5]
Dinh, Tien-Cuong; Nguyˆ en, Viˆ et-Anh; Vu, Duc-Viet: Super-potentials, densities of currents and num- ber of periodic points for holomorphic maps. Adv. Math. 331 (2018), 874–907
work page 2018
-
[6]
Dinh, Tien-Cuong; Nguyˆ en, Viˆ et-Anh; Truong, T uyen Trung: Equidistribution for meromorphic maps with dominant topological degree. Indiana Univ. Math. J. 64 (2015), no. 6, 1805–1828
work page 2015
Show all 26 references
-
[7]
Dinh, Tien-Cuong; Nguyˆ en, Viˆ et-Anh; Truong, T uyen Trung: Growth of the number of periodic points for meromorphic maps. Bull. Lond. Math. Soc. 49 (2017), no. 6, 947–964
2017
-
[8]
Dinh, Tien-Cuong; Sibony , Nessim: Regularization of cu rrents and entropy .Ann. Sci. ´Ecole Norm. Sup. (4) 37 (2004), no
2004
-
[9]
Dinh, Tien-Cuong; Sibony , Nessim: Equidistribution of saddle periodic points for H´ enon-type auto- morphisms of Ck. Math. Ann. 366 (2016), no. 3-4, 1207–1251
2016
-
[10]
Dinh, Tien-Cuong; Sibony , Nessim: Density of positive closed currents, a theory of non-generic inter- sections. J. Algebraic Geom. 27 (2018), 497–551
2018
-
[11]
Dinh, Tien-Cuong; Sibony , Nessim: Unique ergodicity f or foliations in P2 with an invariant curve. Invent. Math. 211 (2018), no. 1, 1–38
2018
-
[12]
Second edition
Fulton, William: Intersection theory . Second edition . Ergebnisse der Mathematik und ihrer Gren- zgebiete. 3. Folge. A Series of Modern Surveys in Mathematic s [Results in Mathematics and Related Areas. 3rd Series. A Series of Modern Surveys in Mathematics ], vol. 2. Springer...
1998
-
[13]
Pure and Applied Mathematics
Griffiths, Phillip; Harris, Joseph: Principles of alge braic geometry . Pure and Applied Mathematics. Wiley-Interscience [John Wiley & Sons], New York, 1978. xii+813 pp
1978
-
[14]
Several complex variables (Proc
Harvey , Reese: Holomorphic chains and their boundarie s. Several complex variables (Proc. Sympos. Pure Math., Vol. XXX, Part 1, Williams Coll., Williamstown, M ass., 1975), pp. 309–382. Amer . Math. Soc., Providence, R. I., 1977
1975
-
[15]
Internat
Kaufmann, Lucas: Self-intersection of foliation cycl es on complex manifolds. Internat. J. Math. 28 (2017), no. 8, 1750054, 18 pp
2017
-
[16]
Kaufmann, Lucas; Vu, Duc-Viet: Density and intersecti on of p1, 1q-currents. J. Funct. Anal. 277 (2019), no. 2, 392–417
2019
-
[17]
International Symposium in Memory of Hua Loo Keng, Vol
Kiselman, Christer O.: Tangents of plurisubharmonic f unctions. International Symposium in Memory of Hua Loo Keng, Vol. II (Beijing, 1988), 157–167, Springer , Berlin,1991
1988
-
[18]
(French) Bull
Lelong, Pierre: Int´ egration sur un ensemble analytiq ue complexe. (French) Bull. Soc. Math. France 85 (1957), 239–262
1957
-
[19]
(French)Gordon & Breach, Paris-London-New York (Distributed by Dunod ´ editeur , Paris) 1968 ix+79 pp
Lelong, Pierre: Fonctions plurisousharmoniques et fo rmes diff´ erentielles positives. (French)Gordon & Breach, Paris-London-New York (Distributed by Dunod ´ editeur , Paris) 1968 ix+79 pp
1968
-
[21]
arXiv:2501.02150 [math.CV ,math.AG, math.DG, math.DS] , 78 pages
Nguyˆ en, Viˆ et-Anh: The generalized Lelong numbers an d intersection theory . arXiv:2501.02150 [math.CV ,math.AG, math.DG, math.DS] , 78 pages
-
[23]
Siu, Yum Tong: Analyticity of sets associated to Lelong numbers and the extension of closed positive currents. Invent. Math. 27 (1974), 53–156
1974
-
[24]
(French) [Extension of closed, positive currents of finite mass] Invent
Skoda, Henri: Prolongement des courants, positifs, fe rm´ es de masse finie. (French) [Extension of closed, positive currents of finite mass] Invent. Math. 66 (1982), no. 3, 361–376. UNIQUENESS OF T ANGENT CURRENTS 33
1982
-
[25]
Thie, Paul R.: The Lelong number of a point of a complex an alytic set. Math. Ann. 172 (1967), 269–312
1967
-
[26]
Vu, Duc-Viet: Densities of currents on non-K¨ ahler manifolds. Int. Math. Res. Not. IMRN. 2021, no. 17, 13282–13304. UNIVERSIT ´E DE LILLE , L ABORATOIRE DE MATH ´EMATIQUES PAUL PAINLEV ´E, CNRS U.M.R. 8524, 59655 VILLENEUVE D ’ASCQ CEDEX , F RANCE . AND VIETNAM INSTITUTE FOR ...
2021
Reviewed August 8, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.