Relating Different Definitions of Linear Series on Tropical Curves

Eric Burkholder

REVIEW 2 major objections 4 minor 13 references

Relating Different Definitions of Linear Series on Tropical Curves

T0 review · 2 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Structured tropical linear series equal combinatorial limit series

desk verdict A genuinely useful unification of tropical linear series definitions, with one definitional mismatch and one missing proof that are fixable; worth serious refereeing. read the letter →

arxiv 2506.16333 v3 pith:3REU64DI submitted 2025-06-19 math.AG math.CO

Eric Burkholder This is my paper · ORCID

classification math.AGmath.CO MSC 14T1014H51

keywords tropicallinearseriesstronglyrecursivecombinatoriallimitpermutationarraysslopestructuresmetricgraphssubmodulesrank

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 attempts to settle whether two recently proposed definitions of linear series on tropical curves describe the same objects. It proves that every strongly recursive tropical linear series of rank $r$ is a combinatorial limit linear series of rank $r$, and then proves the sharper equivalence: a tropical linear series is structured (all of its functions fit inside a fixed slope structure) if and only if it is a combinatorial limit linear series. That matters because the two definitions encode different intuitions, one built from tropical vector-space-like closure and recursive containment, the other from local slope and permutation data at every point; matching them lets results proved in either language carry over. The paper also shows the converse of its main theorem fails for strongly recursive series in ranks $r\ge 3$, so the coincidence is not unconditional: the structured condition is exactly the extra hypothesis that closes the gap.

What carries the argument

The load-bearing object is the local array of a tropical linear series at a point $v$ of the metric graph: the subset of $[r]^d$ recording, for every tangent direction at $v$, which of the $r+1$ slope values is attained by some function in the series. Four closure properties, closure under coordinatewise minima (P1), realization of every slope at every tangent vector (P2), tropical dependence of any $r+2$ elements (P3), and the recursive containment condition (P4), force the array to be the redundant closure of a unique rank-$r$, dimension-$d$ permutation array; here a permutation array is a minimal totally rankable dot array with no redundant dotted positions, and its redundant closure adds all points obtainable as coordinatewise minima of its entries. That description is what lets the paper attach to the graph an $r$-slope structure whose pointwise data are permutation arrays, turning the local data into the admissible-submodule form required by the combinatorial limit definition. The limit argument of Section 4 then supplies the Baker-Norine and local rank properties for effective divisors of degree $r$, completing the bridge between the two definitions.

What would settle it

Inspect the paper's own Figure 5 example at an interior point of an interval: the local array has three dotted points in a $2\times 2$ square, with the corner $(0,0)$ equal to the coordinatewise minimum of the other two. If that corner counts as an extra filled-in point and the slope-structure definition requires the stricter array with no such points, then Corollary 3.13's constructed structure is not a valid slope structure, and Theorem 1.1 is not established by the given proof.

Watch

Extended reading notes

Core claim

The central claim is that the recursive definition of a tropical linear series, augmented by the structured condition that the whole series lies inside a slope structure, is identical to the combinatorial limit definition. Theorem 1.1 proves that any rank-$r$ strongly recursive tropical linear series is a combinatorial limit linear series of rank $r$, and Theorem 5.2 upgrades this to an equivalence: structured tropical linear series are exactly combinatorial limit linear series, so the local-rank condition in the original combinatorial definition can be dropped. The proof route is local: at each point of the metric graph, the series determines a local array of slope-index vectors, and the paper proves this array is the redundant closure of a unique permutation array (Theorem 3.11), which yields the slope structure needed for admissibility. A separate limit argument (Theorem 4.4) produces, for every effective divisor $E$ of degree $r$, a function satisfying both the Baker-Norine rank property and the local rank property for $E$. The converse of Theorem 1.1 fails for ranks $r\ge 3$, and the paper isolates condition (3) of the recursive definition as the obstruction.

Load-bearing premise

The proof assumes that filling in all the extra points obtained by combining entries of a minimal local pattern still counts as the pattern required by the slope structure; if the stricter reading of the definition is used, the constructed slope structure may be invalid and the main theorem is not proved as written.

Editorial extensions

If this is right

Every strongly recursive tropical linear series of rank $r$ automatically satisfies the admissibility conditions of a combinatorial limit linear series of rank $r$.
A tropical linear series is structured exactly when it is a combinatorial limit linear series, giving a shorter definition that drops the separate local-rank requirement.
On the interval and loop metric graphs, and in ranks $0,1,2$ on any metric graph, the tropical-linear-series and combinatorial-limit definitions coincide.
The converse direction fails in ranks $r\ge 3$: there exist combinatorial limit linear series that are not strongly recursive.
Some permutation arrays, including a rank-$3$ array in dimension $4$, cannot be local arrays of strongly recursive tropical linear series, so the local-array description does not extend to all permutation arrays.

Reading between the lines

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

Beyond the paper: if every tropical linear series turns out to be structured, then the two definitions coincide on all tropical linear series; the paper's own open Question 5.3 is the precise test of this.
Beyond the paper: the rank-3 counterexample suggests a general matroid-theoretic obstruction to strong recursion; one could check whether the Levi intersection property characterizes exactly when the recursive condition (3) survives.
Beyond the paper: the rank-2 realizability checks raise the possibility that all rank-2 combinatorial limit linear series are strongly recursive, which would resolve Question 5.11 with a constructive proof for all dimensions.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

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

Read this one if you work on tropical linear series. Burkholder proves that strongly recursive tropical linear series are combinatorial limit linear series (Theorem 1.1), gives a cleaner characterization via slope structures, and constructs counterexamples to the converse for rank at least 3 using Vámos matroids. That is new content, not a survey, and the main line of argument is coherent. The local-array technology in Section 3 is the right tool, and Theorem 3.11—local arrays are redundant closures of permutation arrays—is the key insight.

Where the paper gets soft:

1) Theorem 5.2, the advertised equivalence 'structured iff combinatorial limit linear series,' is asserted after a two-sentence remark that the results 'extend.' That is not a proof. The only-if direction is essentially definitional, but the if direction is exactly the substance of Theorem 1.1 plus a weakened recursion assumption, and the reader is left to fill in an induction that is not written. This needs a real proof or an explicit reduction.

2) Definition 2.14 and Corollary 3.13 do not agree. Definition 2.14(3) asks for a permutation array at each vertex in the strict EL00 sense, no redundant points dotted. Theorem 3.11 shows local arrays are redundant closures, which by definition include redundant points (Example 2.4 interior points already show (0,0) in the local array). Corollary 3.13 calls the local array itself a permutation array. If you take the strict permutation array inside it, Σ⊆R(D,S) fails because functions at redundant dots are excluded. If you loosen the definition to allow redundant points, the slope structure in Corollary 3.13 works and Theorem 1.1 survives, but as written the paper is internally inconsistent. This is repairable—define slope structures via rank arrays or redundant closures—but it is load-bearing for the proof of Theorem 1.1, not cosmetic.

Minor points: the computer checks in Section 6 are reported without code or certificates; for finite checks this is a minor reproducibility issue. Some dependence on [CDI+] and [Wan24] is heavier than the text admits; those are the right citations, but the reader needs them at hand.

Bottom line: the central inclusion is likely true and the framework is a genuine advance; the paper needs a fix to the slope-structure definition and a real proof of Theorem 5.2 before it is accepted. I would send it to a serious referee—ideally one who works on matroids or tropical linear series—rather than desk-reject. It is well above threshold for refereeing, but not ready as is.

Referee Report

2 major / 4 minor

Summary. The paper compares several recent definitions of linear series on tropical curves: strongly recursive tropical linear series (Farkas–Jensen–Payne), tropical linear series in the sense of Chang et al., and combinatorial limit linear series (Amini–Gierczak). The main result, Theorem 1.1, asserts that every strongly recursive tropical linear series of rank r is a combinatorial limit linear series of rank r. The proof proceeds by studying the local arrays of strongly recursive series, showing these arrays are redundant closures of permutation arrays, and then constructing a slope structure and verifying the Baker–Norine and local rank properties. Section 5 extends the main theorem to structured tropical linear series, yielding Theorem 5.2, an equivalence between structured tropical linear series and combinatorial limit linear series. The paper also gives counterexamples to the converse of Theorem 1.1 for rank at least 3, discusses when permutation arrays are realizable as local arrays, and proves low-rank and low-dimensional equivalence results.

Significance. If Theorems 1.1 and 5.2 are correct, the paper achieves a genuine unification of two major recent frameworks for tropical linear series: the recursive definition of Farkas–Jensen–Payne and the slope-structure definition of Amini–Gierczak. The paper is careful and well illustrated, with useful examples (Figures 5 and 6), a clear account of permutation arrays, and explicit low-rank results. It also provides concrete counterexamples, drawn from the Vámos matroid and from Billey–Vakil, showing that certain natural converse statements fail. The machine-checked combinatorial assertions in Section 6 are a further strength. However, the current proof of Theorem 1.1 contains a definitional mismatch between the slope structure in Definition 2.14 and the local arrays produced in Corollary 3.13, and this mismatch is load-bearing for the main theorem.

major comments (2)

[Section 3, Corollary 3.13; Definition 2.14] The r-slope structure S in Corollary 3.13 is not well-defined as written. Definition 2.14(3) requires each P_v to be a permutation array in the strict sense of Subsection 2.3, that is, a totally rankable array with no redundant points dotted. Theorem 3.11, however, proves only that the local array of a strongly recursive tropical linear series is the redundant closure of a permutation array, and Example 3.1 and Figure 5 show that this closure can contain redundant points, e.g. the point (0,0) at the valence-2 points in Example 2.4. Corollary 3.13 then sets P_v equal to the local array and asserts that it is a permutation array; under Definition 2.14 this is false. If one instead takes P_v to be the minimal permutation array inside the local array, the compatibility condition ∂_v(f) ∈ P_v excludes functions that realize redundant local data, so the inclusion Σ ⊆ R(D,S) fails. Thus the proof of Theorem 1.1 has no valid r-slope structure as currently defined. This is repairable by changing the slope-structure definition to use rank arrays or redundant closures rather than strict permutation arrays, but as written Definition 2.14 and Corollary 3.13 are inconsistent.
[Section 5.1, Theorem 5.2] The advertised equivalence is asserted without a proof. The sentence "The results of Section 4 and the proof of Theorem 1.1 extend also to structured tropical linear series" is not a proof of Theorem 5.2. The nontrivial direction, that every structured tropical linear series is a combinatorial limit linear series, requires verifying the admissibility condition of Definition 2.16, in particular the existence, for every effective divisor E of degree r, of a function satisfying both (BNRP) and (LRP). While the induction in Theorem 4.4 likely adapts, the base case uses condition (1) of Definition 2.2 and the induction step uses the slope structure, and these points should be checked explicitly. As written, the main characterization rests on an unproved assertion.

minor comments (4)

[Abstract] The last sentence of the abstract is duplicated: "Finally, address the realizability of permutation arrays as local arrays of linear series on tropical curves" appears twice.
[Section 5.1, after Theorem 5.2] Remark 5.1 says that a combinatorial limit linear series is by definition a structured tropical linear series. This is true only after checking that conditions (1) and (2) of Definition 2.2 hold for an admissible submodule; the implication should be stated explicitly rather than left to the reader.
[Section 6.1, Proposition 6.1] The sentence "Consider the rank rd tropical linear series R(rd·v)" appears to contain a typo or a gap: R(rd·v) has rank rd, not r, and the assertion that any r+1 of its elements are contained in a strongly recursive tropical linear subseries of rank r does not follow directly from condition (3) of Definition 2.3, which applies to sets of size equal to the rank. Please clarify the intended statement and argument.
[Section 4, proof of Theorem 4.4] The compactness argument extracts a convergent subsequence but does not explicitly justify that the needed limits of slopes at every tangent vector exist. Since Lemma 4.2 requires the existence of such limits, a brief diagonal argument over the finite slope data, or an explicit statement that the slopes are bounded integers and stabilize on a further subsequence, would make the proof fully rigorous.

Circularity Check

0 steps flagged · score 0.0 of 10

No circular derivation: Theorem 1.1 rests on new local-array arguments; the Definition 2.14 / Corollary 3.13 tension is a correctness gap, not a circular reduction.

full rationale

The paper's central claim, that strongly recursive tropical linear series are combinatorial limit linear series, is not derived from its own conclusion. Theorem 1.1 is built from genuinely new intermediate results: Section 3 proves that local arrays of strongly recursive series are totally rankable and in fact redundant closures of permutation arrays, and Section 4 proves the required local-rank/Baker–Norine properties using limit arguments. No fitted parameter is renamed as a prediction, and no input object is defined in terms of the target theorem. The cited results from [EL00], [AG22], [FJP25], [JP22], and [CDI+] are supporting lemmas and external definitions; none is authored by Burkholder, and none assumes Theorem 1.1 or the converse as a hypothesis. The one passage that might superficially look circular is Corollary 3.13, where the local array is described as 'a permutation array' after Theorem 3.11 proved it is the redundant closure of a permutation array, while Definitions 2.9–2.10 and Lemma 2.10 reserve the term 'permutation array' for arrays with no redundant points dotted. That is an internal inconsistency or equivocation that can invalidate the proof as written, but it is a correctness/definitional gap rather than a circular derivation of the theorem from its inputs. Hence the circularity score is 0, with the flagged issue belonging to mathematical validity rather than circularity.

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

No numerical parameters are fitted and no new physical entities are introduced. The paper's results rest on previously established lemmas from [JP22], [FJP25], [AG22], [EL00], and [GK08], plus standard compactness and combinatorial facts; these are listed as axioms.

assumptions (6)

domain assumption Finitely generated tropical submodules of R(D) are closed in the sup norm topology.
Invoked in the proofs of Theorem 4.4 and Theorem 1.1; cited to [AG22, Proposition 5.5].
domain assumption For a tropical linear series of rank r, the set of slopes along each tangent vector has exactly r+1 elements.
Lemma 2.5, cited to [FJP25, Lemma 6.6]; used throughout Section 3 to define local arrays.
domain assumption Tropical dependence among functions is preserved by the generated tropical submodule.
Lemma 2.1, cited to [JP22, Lemma 2.4]; used to reduce dependence checks to generator sets.
standard math The characterization of totally rankable dot arrays given by Eriksson and Linusson.
Lemma 2.8, cited to [EL00, Theorem 3.2]; used in Lemma 3.10 and Lemma 5.6 to prove local arrays are totally rankable.
standard math For a permutation array P, the set P is a graded poset and Lemma 2.13 holds.
Lemmas 2.12 and 2.13, cited to [AG22]; used in Theorem 4.1 and Theorem 4.4.
standard math The normalized slice {f in R(D) : f(v0)=0} is compact.
Used in the limit argument in Theorem 4.4; cited to [GK08].

how reviews work

0 comments

Cite this review

Pith. "Pith review of Relating Different Definitions of Linear Series on Tropical Curves." pith.science (2026). https://pith.science/paper/3REU64DI

@misc{pith2026250616333,
  author       = {Pith},
  title        = {Pith review of: Relating Different Definitions of Linear Series on Tropical Curves},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3REU64DI}},
  note         = {Machine review of arXiv:2506.16333}
}

read the original abstract

We investigate relationships among several recent notions of linear series on tropical curves, including tropical linear series, strongly recursive tropical linear series of Farkas, Jensen, and Payne, and combinatorial limit linear series of Amini and Gierczak. We introduce structured and locally weakly recursive tropical linear series, showing that every locally weakly recursive tropical linear series is a combinatorial limit linear series. Consequently, every strongly recursive tropical linear series is a combinatorial limit linear series. We also obtain an equivalent characterization of combinatorial limit linear series. We establish additional extensions of these results and construct counterexamples showing that the converse implications fail for strongly recursive tropical linear series. Finally, we investigate the extent to which permutation arrays can arise as local combinatorial data of tropical linear series.

Figures

Figures reproduced from arXiv: 2506.16333 by the authors.

**Figure 5.** For all points w ∈ (v, x) ∪ (x, u), there are functions in Σ with slope sL[0] on the left tangent vector and slope sR[1] on the right tangent vector, functions with slope sL[0] on the left tangent vector and slope sR[0] on the right tangent vector, and functions with slope sL[1] on the left tangent vector and slope sR[0] on the right tangent vector. Therefore the local array of these points is given as the left dot … view at source ↗

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

13 extracted references · 8 canonical work pages

[1]

Limit linear series: combinatorial theory

Omid Amini and Lucas Gierczak. Limit linear series: combinatorial theory. Preprint arXiv:2209.15613 , 2022

arXiv 2022
[2]

Combinatorial flag arrangements

Omid Amini and Lucas Gierczak. Combinatorial flag arrangements. Preprint arXiv:2404.01971 , 2024

work page Pith review arXiv 2024
[3]

R iemann-- R och and A bel-- J acobi theory on a finite graph

Matthew Baker and Serguei Norine. R iemann-- R och and A bel-- J acobi theory on a finite graph. Advances in Mathematics , 215(2):766--788, 2007

2007
[4]

Intersections of S chubert varieties and other permutation array schemes

Sara Billey and Ravi Vakil. Intersections of S chubert varieties and other permutation array schemes. Algorithms in algebraic geometry , pages 21--54, 2008

work page 2008
[5]

Tropical linear series and matroids

Chih-Wei Chang, Matthew Dupraz, Hernan Iriarte, David Jensen, Dagan Karp, Sam Payne, and Jidong Wang. Tropical linear series and matroids. Preprint arXiv:2508.20062

arXiv
[6]

A tropical proof of the B rill-- N oether theorem

Filip Cools, Jan Draisma, Sam Payne, and Elina Robeva. A tropical proof of the B rill-- N oether theorem. Advances in Mathematics , 230(2):759--776, 2012

work page 2012
[7]

A combinatorial theory of higher-dimensional permutation arrays

Kimmo Eriksson and Svante Linusson. A combinatorial theory of higher-dimensional permutation arrays. Advances in Applied Mathematics , 25(2):194--211, 2000

work page 2000
[8]

The K odaira dimensions of M _ 22 and M _ 23

Gavril Farkas, David Jensen, and Sam Payne. The K odaira dimensions of M _ 22 and M _ 23 . Cambridge J. Math. , 13(3):431--607, 2025

work page 2025

Show all 13 references

[9]

A R iemann-- R och theorem in tropical geometry

Andreas Gathmann and Michael Kerber. A R iemann-- R och theorem in tropical geometry. Mathematische Zeitschrift , 259:217--230, 2008

2008
[10]

Tropical linear series and tropical independence

David Jensen and Sam Payne. Tropical linear series and tropical independence. Preprint arXiv:2209.15478 , 2022

2022 arXiv
[11]

Idempotent analysis, tropical convexity and reduced divisors

Ye Luo. Idempotent analysis, tropical convexity and reduced divisors. Preprint arXiv:1808.01987 , 2018

2018 arXiv
[12]

Tropical curves, their J acobians and theta functions

Grigory Mikhalkin and Ilia Zharkov. Tropical curves, their J acobians and theta functions. Contemporary Mathematics , 465:203--230, 2008

2008
[13]

Lorentzian polynomials and the incidence geometry of tropical linear spaces

Jidong Wang. Lorentzian polynomials and the incidence geometry of tropical linear spaces. Preprint arXiv:2412.12059 , 2024

2024

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [1]

Limit linear series: combinatorial theory

Omid Amini and Lucas Gierczak. Limit linear series: combinatorial theory. Preprint arXiv:2209.15613 , 2022

arXiv 2022

[2] [2]

Combinatorial flag arrangements

Omid Amini and Lucas Gierczak. Combinatorial flag arrangements. Preprint arXiv:2404.01971 , 2024

work page Pith review arXiv 2024

[3] [3]

R iemann-- R och and A bel-- J acobi theory on a finite graph

Matthew Baker and Serguei Norine. R iemann-- R och and A bel-- J acobi theory on a finite graph. Advances in Mathematics , 215(2):766--788, 2007

2007

[4] [4]

Intersections of S chubert varieties and other permutation array schemes

Sara Billey and Ravi Vakil. Intersections of S chubert varieties and other permutation array schemes. Algorithms in algebraic geometry , pages 21--54, 2008

work page 2008

[5] [5]

Tropical linear series and matroids

Chih-Wei Chang, Matthew Dupraz, Hernan Iriarte, David Jensen, Dagan Karp, Sam Payne, and Jidong Wang. Tropical linear series and matroids. Preprint arXiv:2508.20062

arXiv

[6] [6]

A tropical proof of the B rill-- N oether theorem

Filip Cools, Jan Draisma, Sam Payne, and Elina Robeva. A tropical proof of the B rill-- N oether theorem. Advances in Mathematics , 230(2):759--776, 2012

work page 2012

[7] [7]

A combinatorial theory of higher-dimensional permutation arrays

Kimmo Eriksson and Svante Linusson. A combinatorial theory of higher-dimensional permutation arrays. Advances in Applied Mathematics , 25(2):194--211, 2000

work page 2000

[8] [8]

The K odaira dimensions of M _ 22 and M _ 23

Gavril Farkas, David Jensen, and Sam Payne. The K odaira dimensions of M _ 22 and M _ 23 . Cambridge J. Math. , 13(3):431--607, 2025

work page 2025