How statement and proof provenance work
The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.
- Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
- AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
- AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.
These labels describe origin, not correctness: citations and verification chips remain separate evidence.
Kashiwara's singular parabolic kernel filtration
Statement
Let be antidominant, let and assume its full zero-root subsystem is generated by . For each , write for the simple parabolic projection. Then there exists a finite sequence such that for every coherent -module the following are equivalent:
- ;
- there is a finite coherent filtration with for every .
The sequence is one choice, not asserted to be unique; zero successive quotients are permitted.
Facts & Assumptions
Given: Antidominant , its zero simple-root set , the stated zero-root subsystem condition, and coherent .
A reduced expression for gives a fixed coherent chain , where , is generated by finitely many global differential sections, and for coherent partial-TDO modules . This includes the empty-word case (A fixed singular Verma chain with globally generated final bundle).
The simple-parabolic direct-image functors have no positive higher images on coherent modules and are exact there; the displayed quotients in [F1] lie in their respective kernels (The singular simple-parabolic kernel on the flag variety).
For antidominant , global sections is exact (The singular antidominant Serre quotient).
The regular antidominant equivalence applies to all quasicoherent modules, restricts to coherent modules, and expresses any coherent module as a quotient of a finite sum of its TDO by finitely many global sections (Regular antidominant Beilinson–Bernstein localization). Tensoring by an integral line bundle is the TDO-shifting equivalence (PBW and integral line-bundle twisting for flag TDOs).
Proof
Choose once and for all the reduced expression and its simple-root sequence from [F1]. The explicit proof there constructs the Verma embeddings, their rational simple-parabolic descent, and the central-character projector proving global generation of the final term. Thus this finite chain is an actual local supplier, independent of ; it is not an unproved instruction to filter a localization functor. By [F2], each quotient belongs to .
Suppose the asserted filtration of exists. Every factor satisfies . Applying exactness [F3] to gives . Starting with proves .
Now suppose . The weight is regular antidominant. As proved by the simple-reflection and dual-positive-root calculation in [F1], for every positive root , is a positive integer: subtracting it from a nonintegral complex pairing leaves a nonintegral pairing, while subtracting it from a permitted integral pairing, which is nonpositive, gives a strictly negative integer. Thus no positive integral or zero pairing occurs. By the integral twisting equivalence and [F4], is generated by finitely many global differential sections over . Tensor back with ; the twist bimodule identity gives a surjection for some finite , with as in [F1]. Locally this identity is the equality of line-twisted differential operators from the PBW/twist construction in [F4].
Put and . The submodule is the image of the global-section evaluation map and is quasi-coherent; the local filtered-noetherian PBW argument for a TDO makes it coherent as a submodule of the coherent module . Every component in step 1.3 kills because , hence kills , giving a surjection . The globally generated submodule from [F1] is contained in : each of its global generators is a section of . Therefore the images of in give . Their factors are coherent quotients of . Exactness of the corresponding in [F2] puts each factor in its kernel.
Take the direct sum of this chain times and its image under . This gives the required finite coherent filtration of with the same root sequence; every factor is a coherent quotient of a finite sum of factors from step 2.1, and hence has zero corresponding direct image by [F2]. If the sequence is empty, [F1] gives globally generated, so , and , exactly the length-zero filtration. The argument and step 1.2 prove both directions for all the stated complex parameters, with no unproved singular-kernel or derived-localization premise.
Depends on
- The Axiom of Choice
- Complex semisimple algebraic group, Borel, and flag variety
- Ample translation maps split by central-character separation
- Complex-weight separation of Harish-Chandra characters
- The unit V to Gamma Loc_kappa(V) is an isomorphism for antidominant twists
- The singular simple-parabolic kernel on the flag variety
- A fixed singular Verma chain with globally generated final bundle
- PBW and integral line-bundle twisting for flag TDOs
- Regular antidominant Beilinson–Bernstein localization
- The Borel-Weil theorem
- The singular antidominant Serre quotient
Used by
Nothing in the library uses this result yet.
Dependency tree · two levels
69 results within two dependency steps of this one, each drawn at its shortest distance from it. An arrow runs from a result to what uses it, so the chart reads left to right and ends at this result, which carries a heavier outline. Every node is a link to that result. Click elsewhere on the chart to enlarge it.
Sources
- Masaki Kashiwara, Representation theory and D-modules on flag varieties, Astérisque 173–174 (1989) (standard reference, not scraped)