Surjectivity and flatness over DVR's (after Moret-Bailly)

Benedictus Margaux

REVIEW 2 major objections 3 minor 15 references

Surjectivity and flatness over DVR's (after Moret-Bailly)

T0 review · 2 major / 3 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Flatness over a henselian DVR reduces to point lifting

desk verdict A useful but not-yet-citable write-up of Moret-Bailly's DVR lifting criteria; the main theorem is probably right, but the proof of 2.7(a) leans on a mis-cited embedding step that must be repaired. read the letter →

arxiv 2506.01313 v1 pith:S4MBCWFG submitted 2025-06-02 math.AG

Benedictus Margaux This is my paper

classification math.AG MSC 14B2514E22

keywords flatmorphismssurjectivediscretevaluationringshenseliangenericfibreschematicimagepointliftingDedekindschemes

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 gives a point-lifting criterion for flatness over a discrete valuation ring. Fix a henselian DVR $R$, let $\bar R$ be the integral closure of $R$ in an algebraic closure of its fraction field, and let $\bar k$ be the residue field of $\bar R$, an algebraic closure of the residue field $k$ of $R$. The main proposition says that a locally finite type morphism $f: X \to \mathrm{Spec}(R)$ is faithfully flat when the special fibre $X_k$ is integral and the map $X(\bar R) \to X(\bar k)$ is surjective; and conversely, if $f$ is universally open and surjective, that same map is surjective. The reader should care because flatness, usually checked by homological algebra, is here equivalent to a concrete property about lifting points.

What carries the argument

The load-bearing object is the integral closure $\bar R$ of the henselian DVR $R$ in an algebraic closure of its fraction field; it is a henselian local domain whose residue field $\bar k$ is an algebraic closure of $k$, so the map $X(\bar R) \to X(\bar k)$ compares integral points with points over an algebraically closed residue field. The proof works with the schematic image $\tilde X$ of the generic fibre $X_K$ inside $X$, the smallest closed subscheme through which $X_K$ factors; because $\bar R$ is flat over $R$, flat base change lets the author compare $\tilde X$ with $X$ fibre by fibre. To construct lifts, the argument uses the henselian structure theorem for quasi-finite algebras, which decomposes the local ring of $X$ at a point into a factor finite over $R$ with nonzero residue and a factor whose residue is zero; the finite factor is what produces the desired point over $\bar R$.

What would settle it

Find a henselian DVR $R$ and a locally finite type, universally open, surjective morphism $f: X \to \mathrm{Spec}(R)$ with a point of $X(\bar k)$ that does not lift to $X(\bar R)$; alternatively, find an example where $X_k$ is integral and $X(\bar R) \to X(\bar k)$ is surjective but $f$ is not flat. Testing the simplest non-reduced candidates, such as a nilpotent thickening of a smooth fibre over $R$, would show whether the missing embedding step is essential.

Watch

Extended reading notes

Core claim

The central claim (Proposition 2.7 of the paper) is an equivalence. For a henselian DVR $R$ with fraction field $K$ and residue field $k$, let $\bar R$ be the integral closure of $R$ in an algebraic closure of $K$, and let $\bar k$ be its residue field, which is an algebraic closure of $k$. If $f: X \to \mathrm{Spec}(R)$ is locally of finite type, then: (a) whenever $f$ is universally open and surjective, the induced map $X(\bar R) \to X(\bar k)$ is surjective; and (b) whenever the special fibre $X_k$ is integral and $X(\bar R) \to X(\bar k)$ is surjective, $f$ is faithfully flat, meaning flat and surjective. Thus, for locally finite type schemes over a henselian DVR with integral special fibre, flatness is exactly the statement that every algebraic residue point lifts to a point over the integral closure. The paper also proves a weak surjectivity statement for arbitrary noetherian local domains and a flatness criterion for Dedekind schemes with integral fibres of constant dimension.

Load-bearing premise

The proof of part (a) depends on being able to map the finite local factor produced by the henselian decomposition into the integral closure $\bar R$; the result it cites only identifies residue fields, and that embedding is not guaranteed when $X$ is non-reduced.

Editorial extensions

If this is right

If $X_k$ is integral, flatness of $X \to \mathrm{Spec}(R)$ can be certified by checking the single point map $X(\bar R) \to X(\bar k)$ instead of computing flatness from local algebra.
Universally open surjective morphisms over a henselian DVR have $\bar R$-points above every $\bar k$-point, so their special fibres cannot have missing integral points after passing to the integral closure.
A point after a faithfully flat base change, together with integral fibres of constant dimension, forces faithful flatness for schemes over any Dedekind scheme, and if all fibres are smooth the morphism is smooth.
In the weak surjectivity theorem, a faithfully flat locally finite type morphism over a noetherian local domain always admits a section after base change to a valuation ring dominating the base, and over a DVR this valuation ring can be taken to be a DVR.

Reading between the lines

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

A testable extension is to run the criterion on families of curves or abelian schemes over a henselian DVR: the point map $X(\bar R) \to X(\bar k)$ is often computable by Hensel lifting, so the criterion could serve as a practical flatness test in arithmetic geometry.
The equivalence suggests a higher-rank analogue: for a henselian valuation ring of rank greater than one, a similar lifting property might characterize flatness, although the quasi-finite decomposition used here is special to DVRs.
A natural test case is a non-reduced $X$ with integral special fibre; the paper's proof would need an additional argument there, since the local factor used to lift points is only known to be a local ring, not necessarily embeddable into $\bar R$.

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,

Quick take: this note collects Moret-Bailly's lifting criteria over DVRs in one place, with clean organization and mostly careful proofs. But the proof of Proposition 2.7(a) has a real gap that needs fixing before the note can be used as a reference.

What is actually here: weak surjectivity (1.2), flatness/surjectivity criteria (2.5, 2.7), and a Dedekind-scheme version (2.9). The author states openly that essentially all results are due to Moret-Bailly; the value is in the arrangement, the examples, and the proof details. The schematic-image framework in Section 2 is a sensible way to organize the arguments, and 2.9 is a nice extension of the usual group-scheme statement. The references to EGA, Stacks, and Bourbaki are accurate where I checked.

The main problem: in 2.7(a), after producing the finite local R-algebra B from the henselian structure theorem, the proof says 'B embeds into R by 2.6(a)' and uses that to factor X(B) -> X(k) through X(R). That is a misreading of 2.6(a), which gives a map R' -> k, not R' -> R. To lift a B-point to an R-point you need an R-algebra map B -> R compatible with reductions, and that is not granted. If X is non-reduced, B can be non-reduced and no such embedding exists. This is load-bearing for the surjectivity direction. I think it is repairable—choose a minimal prime of B, use flatness over the DVR to embed the domain quotient into R, then handle the residue field by Galois invariance—but that repair is absent.

Two smaller gaps: in 1.2, the total ring of fractions L is treated as a field; it may be a finite product of fields, which is harmless but should be stated. In 2.2(a), the identification of the schematic image with the topological closure needs quasi-compactness; the paper itself warns about the pathology, so the equivalence as stated needs a small hypothesis or a more careful argument.

Bottom line: the attributed theorems are likely correct, and the note is worth having once the gaps are fixed. I would send it to a serious referee, but not accept it as-is.

Referee Report

2 major / 3 minor

Summary. The paper studies locally finite-type morphisms f:X→S with S a DVR or Dedekind scheme. Section 1 proves a weak surjectivity statement: a faithfully flat f admits a valuation-ring point dominating the base. Section 2 analyzes the schematic closure of the generic fibre and gives criteria, culminating in Proposition 2.7 (attributed to Moret-Bailly), which says for a henselian DVR R with integral closure \tilde R in an algebraic closure of Frac(R): (a) if f is universally open and surjective, then X(\tilde R)→X(k) is surjective; (b) if the special fibre is integral and X(\tilde R)→X(k) is surjective, then f is faithfully flat. Applications to Dedekind schemes and fpqc/fppf covers are given. The paper is written as a note with detailed references to EGA, Stacks, Bourbaki, and related literature.

Significance. The results, if fully proved, provide useful and fairly general point-lifting criteria for flatness over DVRs and Dedekind schemes, and the exposition re-derives Moret-Bailly's theorem rather than assuming it. The paper is carefully attributed and contains several correct and useful reductions, notably the reduction of Proposition 2.7(b) to Corollary 2.5. However, two technical gaps in the proofs of Proposition 1.2 and Proposition 2.7(a) currently prevent the manuscript from being fully rigorous; both appear repairable without changing the statements.

major comments (2)

[§2.7(a)] The proof of (a) claims: 'B embeds into R by 2.6(a), making k a B-algebra, so that the map X(B)→X(k) factors through X(R)→X(k).' This is not justified. Fact 2.6(a) gives only a map R'→k when R' dominates R and R'/R is integral; it does not produce an R-algebra map B→\tilde R. Since \tilde R is a domain, an injective map B→\tilde R cannot exist if B is non-reduced, and B may well be non-reduced when X is. The lifting step is load-bearing because it is exactly what turns a B-point into an \tilde R-point. The gap can be repaired by choosing a prime p of B with p∩R=0 (which exists by lying-over for the finite extension R→B), so that B/p is a finite local R-domain dominating R; embedding Frac(B/p) into an algebraic closure of K gives a local homomorphism B→B/p→\tilde R. One must then also arrange that the induced map on residue fields k→k is the identity, or twist by a Galois automorphism of \tilde R over R; the manuscript does not discuss this. This repair should be written out before the proposition can be considered proved.
[§1.2] The proof defines L as the total ring of fractions of R'⊗_R K and states 'Then L is a finite extension of K'. This is generally false: R'⊗_R K is a finite-dimensional K-algebra, and its total ring of fractions is a finite product of finite field extensions, not a field, and it may have nilpotents if the algebra is non-reduced. The subsequent application of [B:AC1, VI, §1.3, Thm. 3], which concerns valuation rings of a field L, is therefore not legitimate. This affects the construction in Proposition 1.2 and in cases (a)–(c). The fix is standard: choose a minimal prime \mathfrak p of R'⊗_R K and put L := Frac((R'⊗_R K)/\mathfrak p); the composite R'→R'⊗_R K→(R'⊗_R K)/\mathfrak p→L supplies the needed R-algebra homomorphism, and the rest of the proof goes through with this field L. This should be stated explicitly.

minor comments (3)

[§2.2(b)] Near display (2.2.3), 'in particular x=h(s)∈X_s' should read 'x=h(s')∈X_s', since s is a point of S and s' is the chosen point of S'.
[§2.5] In the proof of Corollary 2.5, the reference to (2.1.5) in the sentence 'Since Spec(R')→Spec(R) is flat, (2.1.4) and (2.1.5) show that \tilde X_R(R')≅X_R(R')' is unnecessary; the isomorphism follows from (2.1.4) once one knows that Spec(R') is the schematic image of its generic fibre, which itself follows from flatness of R' over R.
[§2.7(a)] After the repair indicated in Major Comment 1, the phrase 'B embeds into R' should be replaced by 'there is a local R-algebra homomorphism B→\tilde R' (for instance via a quotient by a minimal prime over (0)), since an embedding is not what is needed and need not exist when B is non-reduced.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the derivation is self-contained, anchored in external standard references, with no fitted input renamed as a prediction.

full rationale

The paper's main claims (Proposition 2.7, Corollaries 2.4 and 2.5, Proposition 2.9) are proved from classical external sources — EGA, the Stacks Project, Bourbaki, Görtz–Wedhorn, Demazure–Gabriel — with specific section or tag citations at each step. There is no parameter fitted to a subset of data and then presented as a prediction, and no quantity is defined in terms of the target result. The only attribution to prior work is the acknowledgment that the results are due to Moret-Bailly, but the note re-proves them rather than importing them as black boxes; this gives the derivation independent mathematical content. The sole overlap with the author's own work is the generalization of [AG, B.1] in Proposition 2.9, and [AG] is an external reference by Alsaody and Gille, not a self-citation. A possible gap exists in the proof of Proposition 2.7(a), where the assertion 'B embeds into R by 2.6(a)' is not what 2.6(a) states: 2.6(a) provides a homomorphism from the dominating local ring to the algebraic closure of the residue field, not an embedding into the integral closure R. That is a correctness issue in the proof, not a circular reduction of the theorem to its own assumptions. Accordingly, no circular step is exhibited, and the score is 0.

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

No free parameters and no invented entities: this is a pure mathematics note. The central proofs import standard results from EGA, Stacks, Bourbaki, Goertz-Wedhorn, and Demazure-Gabriel; these are listed as axioms. The principal theorems are attributed to Moret-Bailly and re-proved in the text.

assumptions (6)

standard math EGA IV3 14.5.10: a surjective, universally open, locally of finite type morphism over a noetherian scheme admits a finite surjective base change after which it has sections locally everywhere.
Invoked in 1.1(c) as the engine of Proposition 1.2's construction of the valuation ring B.
standard math Bourbaki AC VI, Section 1.3, Theorem 3: the integral closure of R in a finite extension L is the intersection of the valuation rings of L dominating R.
Used in the proof of 1.2 to extract a valuation ring B from a finite cover S'; the proof applies it to L as if L were a field, which is the gap flagged in the report.
standard math Stacks Tag 04GG(13), extending EGA IV4 18.2.1: a quasi-finite algebra over a henselian local ring splits into a finite factor B and a factor C with trivial closed fibre.
The key input of Proposition 2.7(a)'s lifting argument; the paper's inference that B 'embeds into R' goes beyond what the cited statement shows.
standard math EGA IV4 17.16.2, Grothendieck's quasi-finite section theorem: every fppf cover of a scheme admits an affine quasi-finite fppf refinement with a section over it.
Used in Lemma 2.10 to reduce implication (ii) to implication (iii).
standard math EGA IV3 13.1.3 upper semicontinuity of fibre dimension, and Stacks Tag 0062 on stability of closed sets under specializations.
Used in Lemma 2.2(b) to bound fibre dimensions of ~X from below by dim X_eta.
standard math Flat base change for scheme-theoretic images (GW Lemma 14.6, Stacks Tag 01R8) and EGA IV2 2.8.3, the flat closure of the generic fibre over a regular one-dimensional base.
The backbone of Section 2: ~X is flat over S in 2.1(e) and the formation of ~X commutes with the flat base changes used in 2.2(c) and 2.5.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Surjectivity and flatness over DVR's (after Moret-Bailly)." pith.science (2026). https://pith.science/paper/S4MBCWFG

@misc{pith2026250601313,
  author       = {Pith},
  title        = {Pith review of: Surjectivity and flatness over DVR's (after Moret-Bailly)},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/S4MBCWFG}},
  note         = {Machine review of arXiv:2506.01313}
}

abstract

We study morphisms of schemes $f : X \to S$ which are locally of finite type. We present conditions under which there exists a morphism $g : S'\to X$ of $S$--schemes such that $f \circ g $ is the canonical morphism $S'\to S$. Furthermore, we exhibit situations in which $f$ is flat surjective. Our results are mostly concerned with $S$ being the spectrum of a DVR.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

15 extracted references · 15 canonical work pages

[1]

Alsaody, P

S. Alsaody, P. Gille, Isotopes of octonion algebras, G _2 -Torsors and triality , Adv. Math. 343 (2019), 864--909

work page 2019
[2]

Bosch, W

S. Bosch, W. L\"utkebohmert, M. Raynaud, N\'eron models , Ergebnisse der Mathematik und ihrer Grenzgebiete, Band 21, Springer-Verlag 1990

work page 1990
[3]

Bourbaki, Alg\`ebre commutative , Ch

N. Bourbaki, Alg\`ebre commutative , Ch. 5 \`a 7, Masson, Paris 1985

work page 1985
[4]

8 et 9, Masson, Paris 1983

, Alg\`ebre commutative , Ch. 8 et 9, Masson, Paris 1983

work page 1983
[5]

Demazure and P

M. Demazure and P. Gabriel, Groupes alg\'ebriques , North-Holland (1970)

work page 1970
[6]

G\"ortz and T

U. G\"ortz and T. Wedhorn, Algebraic Geometry I , second edition, Springer Fachmedien Wiesbaden, 2020

work page 2020
[7]

Grothendieck (avec la collaboration de J

A. Grothendieck (avec la collaboration de J. Dieudonn\'e), El\'ements de G\'eom\'etrie Alg\'ebrique IV , Publications math\'ematiques de l'I.H.\'E.S. no. 4, 8, 11, 17, 20, 24, 28 and 32 (1964 - 1967)

work page 1964
[8]

Kunz, On Noetherian rings of characteristic p , Amer

E. Kunz, On Noetherian rings of characteristic p , Amer. J. Math. 98 (1976), 999--1013

work page 1976

Show all 15 references

[9]

Raynaud, Anneaux Locaux Hens\'eliens , Lect

M. Raynaud, Anneaux Locaux Hens\'eliens , Lect. Notes in Math. 169, Springer--Verlag, 1970

1970
[10]

Raynaud,Y

M. Raynaud,Y. Laszlo, Expos\'e I. Anneaux excellents , Travaux de Gabber sur l'uniformisation locale et la cohomologie \'etale des sch\'emas quasi-excellents, Ast\'erisque 363-364 (2014), 1-–19

2014
[11]

Schr\"oer, The p -radical closure of local noetherian rings , J

S. Schr\"oer, The p -radical closure of local noetherian rings , J. Commut. Algebra 12 (2020), 135--151

2020
[12]

Serre, Local Fields , Graduate Texts in Mathematics 67, Springer-Verlag, New York-Berlin, 1979

J.-P. Serre, Local Fields , Graduate Texts in Mathematics 67, Springer-Verlag, New York-Berlin, 1979

1979
[13]

Stacks project, http://stacks.math.columbia.edu/
[14]

Vistoli, Grothendieck topologies, fibered categories and descent theory , Fundamental algebraic geometry, 1–-104, Math

A. Vistoli, Grothendieck topologies, fibered categories and descent theory , Fundamental algebraic geometry, 1–-104, Math. Surveys Monogr., 123, Amer. Math. Soc., Providence, RI, 2005

2005
[15]

Warner, Topological fields , North-Holland Mathematics Studies, 157, Amsterdam, 1989

S. Warner, Topological fields , North-Holland Mathematics Studies, 157, Amsterdam, 1989

1989

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

Alsaody, P

S. Alsaody, P. Gille, Isotopes of octonion algebras, G _2 -Torsors and triality , Adv. Math. 343 (2019), 864--909

work page 2019

[2] [2]

Bosch, W

S. Bosch, W. L\"utkebohmert, M. Raynaud, N\'eron models , Ergebnisse der Mathematik und ihrer Grenzgebiete, Band 21, Springer-Verlag 1990

work page 1990

[3] [3]

Bourbaki, Alg\`ebre commutative , Ch

N. Bourbaki, Alg\`ebre commutative , Ch. 5 \`a 7, Masson, Paris 1985

work page 1985

[4] [4]

8 et 9, Masson, Paris 1983

, Alg\`ebre commutative , Ch. 8 et 9, Masson, Paris 1983

work page 1983

[5] [5]

Demazure and P

M. Demazure and P. Gabriel, Groupes alg\'ebriques , North-Holland (1970)

work page 1970

[6] [6]

G\"ortz and T

U. G\"ortz and T. Wedhorn, Algebraic Geometry I , second edition, Springer Fachmedien Wiesbaden, 2020

work page 2020

[7] [7]

Grothendieck (avec la collaboration de J

A. Grothendieck (avec la collaboration de J. Dieudonn\'e), El\'ements de G\'eom\'etrie Alg\'ebrique IV , Publications math\'ematiques de l'I.H.\'E.S. no. 4, 8, 11, 17, 20, 24, 28 and 32 (1964 - 1967)

work page 1964

[8] [8]

Kunz, On Noetherian rings of characteristic p , Amer

E. Kunz, On Noetherian rings of characteristic p , Amer. J. Math. 98 (1976), 999--1013

work page 1976