Generic ordinarity for abelian coverings of the projective line

R\'egis Blache (LAMIA)

REVIEW 2 major objections 4 minor 20 references

Generic ordinarity for abelian coverings of the projective line

T0 review · 2 major / 4 minor · reviewed 2026-08-01 · deepseek-v4-flash

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

Pith's one-line read The paper establishes that abelian covers of the projective line of order prime to p are generically μ-ordinary in characteristic p, meaning the generic Newton polygon of a Hurwitz space component is the minimal polygon allowed by the small

desk verdict Real content and an honest framework, but Theorem 4.5 never shows the Hasse polynomial H_{γ,p} is not identically zero, so generic tightness of the lower-bound polygon — and with it Theorem 1.1 — rests on a missing nonvanishing argument. read the letter →

arxiv 2607.21033 v1 pith:JALZ6LBV submitted 2026-07-23 math.AG math.NT

Régis Blache (LAMIA) This is my paper

classification math.AGmath.NT MSC 11M3814H

keywords NewtonpolygonsHurwitzspacesabeliancoversμ-ordinaryL-functionscharactersumsShimuravarietiesHassepolynomial

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

The paper establishes that, in characteristic p, abelian covers of the projective line of order prime to p are generically μ-ordinary: the generic Newton polygon for a Hurwitz space component coincides with the minimal polygon of the smallest PEL Shimura variety containing its Torelli image. The engine is a computation of generic Newton polygons for L-functions associated to multiplicative character sums over the projective line. The lower bound is attained away from the zero locus of a Hasse polynomial, which is defined by the monodromy datum. If true, this generalizes known p-rank results to the full Newton polygon and yields new Newton polygons realized by Jacobians.

What carries the argument

The p-signature σ(t)=−1+∑⟨p^t a_k/d⟩ is a periodic function whose values govern the slopes of the μ-ordinary polygon Π(γ,e). The lower bound for the Newton polygon of the multiplicative L-function is proved by expressing its coefficients as finite-field character sums, rewriting them via Gauss sums, and applying Stickelberger's congruence. The Hasse polynomial H_{γ,p} is the product over the vertices of the polygon of leading polynomials built from minimal solutions of the congruence system; its vanishing locus is the complement of the open Newton stratum. The identity that carries the argument is the concatenation of the generic Newton polygons of the multiplicative L-functions giving the N

What would settle it

Compute H_{γ,p} for a small explicit datum (for example, d=5, p=2, N=4) and check whether it is the zero polynomial; if it is, then Theorem 4.5 and hence Theorem 1.1 collapse.

Watch

Extended reading notes

Core claim

The central claim is that for any abelian monodromy datum γ=(G,N,a) with |G| prime to p, the Hurwitz space H(γ) has a dense open subset on which the Newton polygon of the Jacobian is the μ-ordinary polygon Π(γ,e) of the smallest Shimura variety containing the Torelli image. The proof reduces to the cyclic case via a cited result and then studies the family of polynomials g=∏(x−α_i)^{a_i}. Theorem 4.5 identifies the generic Newton polygon of the associated L-functions L(g,χ,T) with Π(γ,e), and shows it is attained exactly when the Hasse polynomial H_{γ,p} is nonzero. The μ-ordinary polygon is determined by the p-signature σ(t) of the monodromy datum, a periodic function of t.

Load-bearing premise

The proof assumes the Hasse polynomial H_{γ,p} is not identically zero; if it were, the open Newton stratum would be empty and the lower-bound polygon would never be attained.

Editorial extensions

If this is right

The Hurwitz space H(γ) intersects the μ-ordinary Newton stratum of the associated Shimura variety in a dense open subset, so a Zariski-generic abelian cover is μ-ordinary.
Combining with known agreement of open Ekedahl–Oort and Newton strata in Shimura varieties, generic abelian covers are also [p]-ordinary.
The result enriches the set of Newton polygons known to be realized by Jacobians and yields new unlikely intersections of the Torelli locus with Newton strata.
The generic Newton polygon for the family of L-functions is now explicit for all d, p, and a, not just under large-characteristic or few-branch-point conditions.

Reading between the lines

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

A testable extension is to compute the Hasse polynomial for small explicit monodromy data to confirm it is not identically zero; if a zero example existed, the generic polygon would be strictly above Π(γ,e) and the main theorem would fail.
The argument relies on a cited reduction from abelian to cyclic covers; if that reduction were incomplete, a direct proof for abelian covers would be needed to close the gap.
The techniques might extend to character sums over higher-dimensional projective spaces, where generic Newton polygons are currently only known in special cases.
The μ-ordinary polygon depends only on the residue of p modulo d, a much weaker p-dependence than in the additive character-sum case, hinting at a structural difference between the two settings.

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

The short version: real content, honest framing, and one load-bearing gap in the proof of Theorem 4.5 that has to be closed before the main theorem stands.

The genuinely new material is the claim that the lower-bound polygon Π(γ,e) — already known from [Dol13, Thm 5.1] except for endpoint coincidence, as the paper itself says — is generically tight, together with the explicit Hasse-polynomial criterion and the transfer from L-functions to arbitrary abelian prime-to-p covers. The machinery in Sections 2 and 3 is mostly in good shape: the counting of A_n is consistent, the Stickelberger valuation argument is coherent, the endpoint equality at n = N−2 is handled, and the base-p^r digit structure (Prop 3.14) with the norm factorization (Prop 4.3) is genuinely elegant. That part deserves real credit.

The problem is exactly where the stress-test puts it. Theorem 4.5 proves a criterion: the polygon is attained at α iff H_{γ,p}(α) ≠ 0. It never proves H_{γ,p} is not the zero polynomial. The last sentence of the proof restates the equivalence; it does not exhibit an α or give an argument that the product of vertex polynomials is not identically zero. Nonemptiness of M_n (Prop 3.13) is a different claim from nonvanishing of the sum over M_n. The vertices sit at n ≤ N−3, where multiple minimal solutions exist and cancellation is possible in principle. If H_{γ,p} ≡ 0 for some γ and p, the lower bound is never attained, the generic Newton polygon is strictly higher, and Theorem 1.1 collapses. The introduction announces that the principal parts are shown to be generically nonzero; Section 4.2 does not deliver that. This is a missing argument rather than a demonstrated error, and my guess is it is patchable — some specialization or monomial-independence argument — but as written it is absent.

Two secondary issues, both minor: the abelian-to-cyclic reduction is cited to [LMS24, §3] rather than restated, and the transfer from L-functions to Jacobians in the introduction is a sketch. A referee will want those expanded, but they are dependencies on published work, not circular steps.

Who this is for: people working on Newton stratifications of Hurwitz spaces or on multiplicative character-sum L-functions. It deserves a serious referee. My recommendation: send it to peer review, with a report that asks for the nonvanishing proof. If that comes out cleanly, the paper is a solid contribution.

Referee Report

2 major / 4 minor

Summary. The paper claims that every abelian cover of P^1 whose Galois group has order prime to p is generically μ-ordinary in characteristic p. After reducing to cyclic covers (by a cited result), it analyzes multiplicative character-sum L-functions L(g,χ;T) for g(x)=∏(x−α_i)^{a_i}. The main technical result, Theorem 4.5, identifies the generic Newton polygon of this family with a combinatorial polygon Π(γ,e) and defines a Hasse polynomial H_{γ,p} whose nonvanishing locus is asserted to be exactly the locus where Π(γ,e) is attained. The paper then concludes Theorem 1.1 by concatenating these polygons for all characters of the cyclic group.

Significance. If the main theorem is fully established, it is a substantial result: it extends known μ-ordinarity statements from cyclic covers with few branch points to arbitrary prime-to-p abelian covers, and it gives new evidence for the intersection of the Torelli locus with Shimura-variety Newton strata. The paper is carefully written and contains several valuable technical contributions: the Stickelberger-based lower bound in Proposition 2.6, the combinatorial study of minimal solutions in Section 3, and the elegant norm factorization of the vertex polynomials in Proposition 4.3. The lower-bound argument appears coherent and is largely self-contained. However, the proof of Theorem 4.5 leaves a load-bearing gap: it never proves that the Hasse polynomial H_{γ,p} is nonzero as a polynomial, which is necessary for the claimed generic attainment.

major comments (2)

[Theorem 4.5 / §4.2] The proof does not establish that H_{γ,p} is a nonzero polynomial. Lemma 4.1 and Prop. 4.3 show only that, for a fixed α, the vertex π of Π(γ,e) is attained iff H^{(π)}_{γ,p^r}(α)≠0; hence Π is attained iff H_{γ,p}(α)≠0. But being the generic Newton polygon requires a nonempty open subset of A^N on which equality holds, i.e. H_{γ,p} not identically zero. The sentence "Guaranteeing that the product over the vertices is non zero..." restates the equivalence; it does not prove existence of such α. Prop. 3.13 gives nonemptiness of M_n, but a nonempty index set does not prevent cancellation modulo p in the sum in Def. 4.2. If H_{γ,p}≡0 for some γ,p, the lower bound Π is never attained and Theorem 1.1 collapses. Please add a nonvanishing proof (e.g., a monomial-order argument or a specialization).
[§1, proof of Theorem 1.1] The proof of Theorem 1.1 relies on a reduction from abelian to cyclic monodromy, cited as [LMS24, Section 3], but no statement of the reduction is given. Since the theorem concerns arbitrary abelian G, the paper should either state the reduction theorem and explain how μ-ordinarity for the associated cyclic covers implies μ-ordinarity for the original abelian family, or give a precise reference with the hypotheses verified.

minor comments (4)

[§2.3, Prop. 2.6] In the displayed equality after the first inequality, the second summation should run from t=0 to r−1, not m−1; as printed the equality is false when m>r.
[Def. 3.3] 'fiwesq−1' should be 'fixes q−1'. Also, 'shift' is better described as the cyclic digit permutation on {0,...,q−2}.
[Throughout] Several typographical errors: 'L-fonction' in §2, 'we now from a theorem' in §2.3, 'bur' and 'lasr sum' in §3.2. These do not affect the mathematics.
[§4.1] The definition of the leading polynomial in Def. 4.2 would be clearer if it explicitly stated that the exponents q−1−u_ij are nonnegative for minimal solutions, using Lemma 3.2.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the generic-polygon computation is derived from congruences and Gauss sums, not from its own conclusion.

full rationale

The central derivation is not circular. The lower bound for the L-function Newton polygon (Prop. 2.6 and Cor. 2.10) is proved directly from Stickelberger valuations and the Ax/Davenport-Hasse expression for coefficients; it does not presuppose the generic polygon. The criterion in Theorem 4.5, 'It is attained if, and only if we have H_{γ,p}(α)≠0,' is obtained in Lemma 4.1 by explicit congruences, and H is then defined as the product of the resulting vertex polynomials. That is a computed criterion, not a fitted parameter renamed as a prediction. The reduction of abelian covers to cyclic covers is cited from [LMS24, Section 3]; this is an external dependency and not a self-citation. The self-citations to [Bla12] and [BF07] concern techniques and context only and are not load-bearing for the main theorem. The real weakness is that the final sentence of the proof of Theorem 4.5, 'Guaranteeing that the product over the vertices is non zero gives a necessary and sufficient condition for the coincidence of the two polygons,' only restates the criterion; the paper does not explicitly prove H_{γ,p} is not identically zero, which is needed to show the lower-bound polygon is actually attained on a nonempty Zariski open set. This is an unproved nonvanishing step, hence a correctness gap, but it is not a circular reduction of the conclusion to its inputs.

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

No fitted constants are introduced; d,N,a,p are inputs. The main unproved inputs are external theorems (specialization, Shimura stratification, LMS24 reduction) plus an asserted concatenation identity and the implicit non-vanishing of the Hasse polynomial. The Hasse polynomial is constructed explicitly, so it is not an unexplained invented entity.

assumptions (6)

standard math For a cyclic cover C: y^d = ∏(x−α_i)^{a_i}, the zeta function factors as L(C;T)=∏_{k=1}^{d−1} L(χ,g^k;T), and the Newton polygon of the Jacobian is the concatenation of the factors' Newton polygons.
Invoked in the Introduction to reduce Theorem 1.1 to Theorem 4.5; standard character-sum factorization, not proved in the paper.
standard math Grothendieck's specialization theorem: for the family of polynomials with distinct roots, a generic Newton polygon exists and is attained on a nonempty Zariski open subset.
Used in Section 4 preamble; cited to [Kat79, Theorem 2.3.1].
domain assumption The Newton polygon stratification of PEL-type Shimura varieties is known, and the μ-ordinary (lowest) polygon is the generic one.
Used to identify the target polygon for the Torelli image; cited to [RR96], [VW13], [Moo04].
domain assumption Arbitrary abelian covers reduce to cyclic covers for Newton polygon purposes, via [LMS24, Section 3].
The proof of Theorem 1.1 says 'we can reduce to the cyclic case from [LMS24, Section 3]'; this is an external result on which the main theorem depends.
ad hoc to paper The concatenation of the generic Newton polygons of the individual multiplicative L-functions is the μ-ordinary polygon for the monodromy datum γ.
Stated in the Introduction ('Since the concatenation ... is the μ-ordinary polynomial for γ') without a detailed proof; it is the bridge from Theorem 4.5 to Theorem 1.1.
standard math Stickelberger's congruence for Gauss sums and the Ax interpolation-polynomial method give the valuations and congruences for L-function coefficients.
Used throughout Sections 2 and 4; cited to [Sti90], [Ax64], [AS14].

how reviews work

0 comments

Cite this review

Pith. "Pith review of Generic ordinarity for abelian coverings of the projective line." pith.science (2026). https://pith.science/paper/JALZ6LBV

@misc{pith2026260721033,
  author       = {Pith},
  title        = {Pith review of: Generic ordinarity for abelian coverings of the projective line},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/JALZ6LBV}},
  note         = {Machine review of arXiv:2607.21033}
}

abstract

We show that abelian coverings of the projective line of order prime to $p$ are generically $\mu$-ordinary in characteristic $p$. The images of the irreducible components of Hurwitz spaces of abelian coverings of the projective line by the Torelli morphism lie in some Shimura varieties. The stratification by Newton polygons of these varieties is known, and we show that the generic Newton polygon for the Hurwitz space coincides with the generic (or $\mu$-ordinary) Newton polygon of the smallest Shimura variety that contains its image. In order to do this, we compute the generic Newton polygons for $L$-functions associated to multiplicative character sums over the projective line.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

20 extracted references · 1 linked inside Pith

[1]

Number Theory 142 (2014), 183--210

Alan Adolphson and Steven Sperber, Hasse invariants and mod $p$ solutions of $A$ -hypergeometric systems , J. Number Theory 142 (2014), 183--210

2014
[2]

James Ax, Zeroes of polynomials over finite fields, Am. J. Math. 86 (1964), 255--261

1964
[3]

Number Theory 123 (2007), no

R \'e gis Blache and \'E ric F \'e rard, Newton stratification for polynomials: the open stratum, J. Number Theory 123 (2007), no. 2, 456--472

2007
[4]

Number Theory 132 (2012), no

R \'e gis Blache, Valuation of exponential sums and the generic first slope for Artin - Schreier curves , J. Number Theory 132 (2012), no. 10, 2336--2352

2012
[5]

Bouw, The $p$ -rank of ramified covers of curves

Irene I. Bouw, The $p$ -rank of ramified covers of curves. , Compos. Math. 126 (2001), no. 3, 295--322

2001
[6]

Reine Angew

Harold Davenport and Helmut Hasse, Die Nullstellen der Kongruenzzetafunktionen in gewissen zyklischen F \"a llen , J. Reine Angew. Math. 172 (1934), 151--182

1934
[7]

John Dollarhide, On the $L$ -function of multiplicative character sums , Trans. Am. Math. Soc. 365 (2013), no. 3, 1637--1668

2013
[8]

Katz, Slope filtration of f -crystals, Journ\'ees de G\'eom\'etrie Alg\'ebrique de Rennes, Ast\'erisque, no

Nicholas M. Katz, Slope filtration of f -crystals, Journ\'ees de G\'eom\'etrie Alg\'ebrique de Rennes, Ast\'erisque, no. 63, Soci\'et\'e math\'ematique de France, 1979, pp. 113--163

1979

Show all 20 references

[9]

10 (2004), no

, Notes on $G_ 2 $ , determinants, and equidistribution , Finite Fields Appl. 10 (2004), no. 2, 221--269

2004
[10]

Number Theory 5 (2019), no

Wanlin Li, Elena Mantovan, Rachel Pries, and Yunqing Tang, Newton polygons arising from special families of cyclic covers of the projective line, Res. Number Theory 5 (2019), no. 1, 31, Id/No 12

2019
[11]

, Newton polygon stratification of the Torelli locus in unitary Shimura varieties , Int. Math. Res. Not. (2022), no. 9, 6464--6511

2022
[12]

Yuxin Lin, Elena Mantovan, and Deepesh Singhal, Abelian covers of $ P ^1$ of $p$ -ordinary Ekedahl - Oort type , Int. Math. Res. Not. (2024), no. 23, 14369--14392

2024
[13]

Hanfeng Li and Hui June Zhu, Zeta functions of totally ramified $p$ -covers of the projective line , Rend. Semin. Mat. Univ. Padova 113 (2005), 203--225

2005
[14]

37, 2004, pp

Ben Moonen, Serre- Tate theory for moduli spaces of PEL type , vol. 37, 2004, pp. 223--269

2004
[15]

, Special subvarieties arising from families of cyclic covers of the projective line, Doc. Math. 15 (2010), 793--819

2010
[16]

Antonine Phigareau, The p-powers dividing certain exponential sums, Finite Fields and Their Applications 117 (2027), 102876

2027
[17]

Rachel Pries, The Torelli locus and Newton polygons , Preprint, arXiv :2509.00998

arXiv
[18]

Rapoport and M

M. Rapoport and M. Richartz, On the classification and specialization of $F$ -isocrystals with additional structure , Compos. Math. 103 (1996), no. 2, 153--181

1996
[19]

Stickelberger, Ueber eine Verallgemeinerung der Kreistheilung

L. Stickelberger, Ueber eine Verallgemeinerung der Kreistheilung . , Math. Ann. 37 (1890), 321--367
[20]

Eva Viehmann and Torsten Wedhorn, Ekedahl- Oort and Newton strata for Shimura varieties of PEL type , Math. Ann. 356 (2013), no. 4, 1493--1550

2013

Pith tools

Reviewed August 1, 2026 · model on record in the stance chip above.

[1] [1]

Number Theory 142 (2014), 183--210

Alan Adolphson and Steven Sperber, Hasse invariants and mod \(p\) solutions of \(A\) -hypergeometric systems , J. Number Theory 142 (2014), 183--210

2014

[2] [2]

James Ax, Zeroes of polynomials over finite fields, Am. J. Math. 86 (1964), 255--261

1964

[3] [3]

Number Theory 123 (2007), no

R \'e gis Blache and \'E ric F \'e rard, Newton stratification for polynomials: the open stratum, J. Number Theory 123 (2007), no. 2, 456--472

2007

[4] [4]

Number Theory 132 (2012), no

R \'e gis Blache, Valuation of exponential sums and the generic first slope for Artin - Schreier curves , J. Number Theory 132 (2012), no. 10, 2336--2352

2012

[5] [5]

Bouw, The \(p\) -rank of ramified covers of curves

Irene I. Bouw, The \(p\) -rank of ramified covers of curves. , Compos. Math. 126 (2001), no. 3, 295--322

2001

[6] [6]

Reine Angew

Harold Davenport and Helmut Hasse, Die Nullstellen der Kongruenzzetafunktionen in gewissen zyklischen F \"a llen , J. Reine Angew. Math. 172 (1934), 151--182

1934

[7] [7]

John Dollarhide, On the \(L\) -function of multiplicative character sums , Trans. Am. Math. Soc. 365 (2013), no. 3, 1637--1668

2013

[8] [8]

Katz, Slope filtration of f -crystals, Journ\'ees de G\'eom\'etrie Alg\'ebrique de Rennes, Ast\'erisque, no

Nicholas M. Katz, Slope filtration of f -crystals, Journ\'ees de G\'eom\'etrie Alg\'ebrique de Rennes, Ast\'erisque, no. 63, Soci\'et\'e math\'ematique de France, 1979, pp. 113--163

1979