A common framework for test ideals, closure operations, and their duals

Closure operations such as tight and integral closure and test ideals have appeared frequently in the study of commutative algebra. This articles serves as a survey of the authors’ prior results connecting closure operations, test ideals, and interior operations via the more general structure of pair operations. Specifically, we describe a duality between closure and interior operations generalizing the duality between tight closure and its test ideal, provide methods for creating pair operations that are compatible with taking quotient modules or submodules, and describe a generalization of core and its dual. Throughout, we discuss how these ideas connect to common constructions in commutative algebra.

💡 Research Summary

**
This paper surveys and unifies a broad collection of closure operations, test ideals, and interior (or “dual”) operations that have appeared throughout commutative algebra, by placing them within the general framework of pair operations. A pair operation p assigns to each inclusion of modules L ⊆ M a submodule p(L,M) ⊆ M, satisfying natural compatibility with isomorphisms. The authors first collect the elementary properties that such operations may enjoy—extensiveness (L ⊆ p(L,M)), intensiveness (p(L,M) ⊆ L), idempotence (p(p(L,M),M)=p(L,M)), order‑preservation on submodules, and various functoriality conditions (surjection‑functorial, co‑functorial, fully functorial). They then show how classical closure operations (integral closure, tight closure, Ratliff‑Rush, a‑tight closure, etc.) fit into this language, and how many “interior” operations (tight interior, homogeneous interior, basic empty interior) are also examples of pair operations, but with the intensiveness condition instead of extensiveness.

A central contribution is the introduction of the “smile dual” operation p↦p⌣. In the setting of a complete local Noetherian ring R with injective hull E of the residue field, the dual is defined by
 p⌣(A,B) = Ann_R((Ann_E A) p E).
When p is a closure operation, p⌣ is an interior operation, and vice‑versa. The authors prove that in Artinian or Noetherian contexts p⌣⌣ = p, and that the dual exchanges the basic properties: extensiveness ↔ intensiveness, idempotence ↔ idempotence, etc. This generalizes the well‑known duality between tight closure and its big test ideal, and shows that many other closure–test‑ideal pairs are instances of the same phenomenon.

The paper also develops systematic ways to modify a given pair operation so that it satisfies residual or hereditary properties. Residuality ensures compatibility with intersections (p(L,M)∩N = p(L∩N,N)), while hereditaryness guarantees compatibility with submodule restriction (if p(L,M)⊆N then p(L,N)=p(L,M)∩N). By applying these constructions, one can produce from any closure operation a “core” operation (the intersection of all reductions) and a dual “hull” operation (the sum of all expansions), extending the classical notions of core(I) and hull(I) to arbitrary Nakayama closures.

Section 2.4 culminates in Theorem 2.4.7, which gives a precise duality between a closure operation cl and its associated test ideal τ(cl): for an ideal I, τ(cl)(I) = Ann_R((Ann_E I) cl E). This recovers the tight‑closure/test‑ideal correspondence and shows that the same formula works for integral closure, basically full closure, and many other examples.

Section 3 provides a comprehensive catalogue of known pair operations, tabulating their properties (extensive, intensive, idempotent, functorial, persistent, etc.) and supplying proofs or counterexamples where appropriate. The authors then illustrate how the theory applies to module closures and trace ideals: given a module M, the trace ideal Tr(M) is the sum of images of all homomorphisms from M to R; the duality framework shows that the trace ideal can be viewed as the interior dual of a suitable closure operation on submodules of M. This perspective unifies several results about cores of ideals, hulls of modules, and test ideals for modules.

Finally, the paper discusses core and hull for arbitrary Nakayama closures, establishing that the core of a submodule (the intersection of all cl‑reductions) and the hull (the sum of all cl‑expansions) are dual under the smile operation. Applications to basically full closure demonstrate that the “full test ideal” is precisely the hull of the zero submodule under the dual interior operation.

Overall, the article presents a coherent, categorical language that captures a wide array of closure‑type constructions in commutative algebra, explains their interrelations via a universal duality, and supplies practical tools (residual/hereditary adjustments, meets/joins, persistence criteria) for building new operations with desired properties. It serves both as a reference guide to existing literature and as a springboard for future research on novel closure operations, their test ideals, and associated core/hull theories.