Published paper
Systèmes holonomes d’équations aux -différences
MSC 2020
33D6032C3839A10
Abstract
The paper develops a holonomic theory for systems of -difference equations parallel to holonomic modules over the Weyl algebra, including canonical transformations, a Bernstein inequality, functorial operations, and applications to -hypergeometric systems.
Role in dependence graphs
Proof-critical source
On Lagrangianity of -supports of holonomic -modules and --modules
This paper is included only for the following marked statements:
- Subsection 1.3 · published pagination pp. 128–129 / author-manuscript fallback pp. 4–5Integral symplectic transformations act by automorphisms of the -difference operator algebra.
- Corollary 2.1.2 · published pagination p. 133 / author-manuscript fallback p. 9Bernstein inequality for nonzero finitely generated modules over the -difference operator algebra.
AI-generated audit
Audit summary
Not a correctness certificate. These reports do not replace expert scrutiny or formal verification.
Exact reviewed source
Official author manuscript · explicit fallback for the inaccessible De Gruyter chapter version of record
Claude Sabbah. Systèmes holonomes d'équations aux -différences. In D-Modules and Microlocal Geometry, edited by Masaki Kashiwara, Teresa Monteiro Fernandes, and Pierre Schapira, De Gruyter, 1993, pages 125-147.
The exact De Gruyter chapter version of record was not publicly downloadable during this audit. Claude Sabbah's official publication page supplies a 24-page author manuscript for the cited chapter, so that manuscript was reviewed as the explicit proof-bearing fallback; this report does not claim inspection of the publisher file or its typography.
Open audited source ↗01Statements6 reported findingsSuspected incorrect statements
The main Section 2 results used by the focal paper - simplicity, Bernstein's inequality, finite generic rank, and the holonomic-lattice construction - have the necessary non-root-of-unity hypothesis and a coherent local statement chain when Theorem 2.7.1 is read in the lattice scope used by its proof. As printed, however, Proposition 1.4.1 is false without that hypothesis, Theorem 2.7.1 omits its lattice qualifier and localization conclusion, and Section 3's infinite-product realization lacks the convergence condition it needs.
Section 2 core chainCorrect with the stated Section 2 hypothesis
The algebraic results cited downstream have the required non-root-of-unity scope
Author-manuscript pages 3 and 8-19 · Proposition 1.2.1, Corollary 2.1.2, Corollary 2.5.2, and Theorem 2.7.1 · published pages 127 and 132-143
Proposition 1.2.1 proves simplicity of when is not a root of unity. At the start of Section 2 the chapter expressly keeps that hypothesis in force, and the commutator step in the Bernstein-inequality proof uses it. The restriction argument then gives finite dimension after extension to , while the proof of Theorem 2.7.1 constructs a holonomic lattice inside a finite lattice that already spans the rational system. These are the source results actually invoked by the focal paper.
Official author manuscript ↗Theorem 2.7.1(2)Suspected incomplete statement · uniquely repairable
The printed statement omits the lattice condition used by its proof
Author-manuscript page 19 · Theorem 2.7.1 and proof · published page 143
Part (2) quantifies over every finite -submodule contained in the rational system, but the proof immediately assumes and ends by proving the same equality for the holonomic submodule it constructs. Without the spanning condition, that conclusion is false for a proper invariant rational subsystem; without the conclusion, the literal statement is vacuous because the zero module is holonomic. The intended and proved formulation is: for every finite lattice with , there is a holonomic with .
Official author manuscript ↗Proposition 1.4.1Suspected incorrect at printed scope · hypothesis missing
The root-of-unity case contradicts the claimed free orbit decomposition
Author-manuscript page 5 · Proposition 1.4.1 and proof · published page 129
The proposition is stated before the chapter imposes the standing assumption that is not a root of unity. Let be a primitive th root of unity and, for , let have basis with indices modulo , , and . This is a nonzero finite-dimensional cyclic module for the relation , but it cannot be isomorphic to the claimed free module for any nonzero finite-dimensional . Equivalently, the proof's generalized eigenspaces cease to give an infinite direct orbit decomposition when the -orbit repeats. Adding “ is not a root of unity” repairs both statement and proof and is consistent with the standing assumption introduced at the start of Section 2.
Official author manuscript ↗Section 3.1 explicit realizationSuspected incorrect at printed scope · convergence hypothesis missing
The infinite product is not meromorphic for the chapter's full range of
Author-manuscript page 20 · formula following Proposition 3.1.1 · published page 144
The chapter defines and immediately calls the resulting expression quasi-meromorphic, while its only standing restriction is that is not a root of unity. For example, when and , the factors do not tend to , so the product cannot converge. The displayed analytic realization is valid after imposing ; other values require a different normalization or a separate analytic treatment. This does not by itself refute the preceding algebraic classification of rank-one rational systems.
Official author manuscript ↗Section 1.3 canonical transformationsCorrect with an important scope boundary
Composition is explicitly only up to powers of
Author-manuscript pages 4-5 · Lemmas 1.3.1-1.3.2 and following discussion · published pages 128-129
An integral symplectic matrix gives an automorphism by sending the coordinates to the associated monomials. The chapter also records that composing two such choices produces the monomials for the product matrix only up to scalar powers of . Thus the source supports individual canonical automorphisms, but a literal action of requires an additional normalization or central-extension argument not asserted here.
Official author manuscript ↗Proposition 1.1.1 and Section 2.5.3Suspected typos
Two local index and open-set errors have unique mechanical corrections
Author-manuscript pages 3 and 16 · Proposition 1.1.1(2) and Section 2.5.3 · published pages 127 and 140
In Proposition 1.1.1(2), the quotient family is indexed by the divisors , so “for every ” must read “for every .” In Section 2.5.3 the localization set is called a Zariski open and is built by inverting denominator translates, so “the open sets of equation ” must read “the open sets .” Both corrections are forced by the definitions and leave the subsequent arguments unchanged.
Official author manuscript ↗02Proofs3 reported findingsFurther proof verification needed
The elementary quantum-torus calculations, Bernstein-inequality induction, restriction argument, and finite-rank deduction are coherent at source level. Several broader transfers from Weyl-algebra and -module theory are deliberately compressed; most notably, Theorem 2.6.2 relies on substantive specialization lemmas that are asserted only by analogy and are not proved or cited in their -difference form.
Theorem 2.1.1 through Corollary 2.5.2Checked at source level
The cited Bernstein and finite-rank chain closes locally
Author-manuscript pages 8-16 · Sections 2.1-2.5 · published pages 132-140
The Bernstein proof lowers the chosen operator degree by commuting with a coordinate, using that distinct powers of a non-root-of-unity cannot coincide. Holonomicity is then preserved by the explicit codimension-one restriction complex, and Proposition 2.5.1 applies that restriction after adjoining an independent generic coordinate. Iterating gives Corollary 2.5.2, the finite-dimensional generic fiber used downstream. The module and field changes in these steps retain the stated hypotheses.
Official author manuscript ↗Theorem 2.6.2Proof not independently verified
The specialization proof imports two unproved -difference analogues
Author-manuscript pages 18-19 · Section 2.6, especially Theorem 2.6.2 · published pages 142-143
After establishing an elementwise Bernstein relation, the chapter asserts that every specializable module has a resolution by elementary specializable modules and that every morphism between specializable modules is strict for the -filtration. Those are the steps needed to make the associated graded module holonomic, but neither is proved for the noncommutative -difference algebra; the cited references treat the -module analogue. The conclusion may be true, but this chapter alone does not verify this proof transfer.
Official author manuscript ↗Ordinary-audit verification boundaryNot independently rederived
The long homological transfers were not independently reconstructed
Author-manuscript pages 6-24 · Theorem 1.5.3 and later Weyl-algebra analogies · published pages 130-147
The chapter repeatedly adapts standard-basis, homological-dimension, direct-image, and hypergeometric-module arguments from the Weyl-algebra literature, often by stating the quantum-commutative replacement rather than reproducing every lemma. The complete manuscript was inspected and the focal dependency chain was traced, but an ordinary audit does not reproduce every one of those long adaptations line by line. This coverage limit is separate from the concrete missing hypotheses and the specific specialization gap recorded above.
Official author manuscript ↗03Novelty0 reported findingsNo novelty concerns found
No non-novelty findings.