What is the role of continuity in continuous linear forms representation?

The recent extensions of domain theory have proved particularly efficient to study lattice-valued maxitive measures, when the target lattice is continuous. Maxitive measures are defined analogously to classical measures with the supremum operation in place of the addition. Building further on the links between domain theory and idempotent analysis highlighted by Lawson (2004), we investigate the concept of domain-valued linear forms on an idempotent (semi)module. In addition to proving representation theorems for continuous linear forms, we address two applications: the idempotent Radon–Nikodym theorem and the idempotent Riesz representation theorem. To unify similar results from different mathematical areas, our analysis is carried out in the general Z framework of domain theory.

💡 Research Summary

The paper investigates the interplay between domain theory and idempotent analysis, focusing on the role of continuity in the representation of linear forms defined on idempotent (semi)modules whose values lie in a continuous lattice. The authors begin by recalling that a maxitive measure replaces the additive operation of classical measure theory with the supremum (join) operation, and that when the target lattice is continuous, many of the classical measure‑theoretic constructions admit natural analogues. Recent advances in domain theory—particularly the development of the Z‑framework, which generalizes continuous lattices to a broader categorical setting—provide the appropriate language for handling such structures.

A central object of study is a linear functional φ : M → L, where M is an idempotent semimodule over a complete idempotent semiring and L is a continuous lattice. The authors define φ to be continuous precisely when it preserves directed suprema, i.e., it is Scott‑continuous. This notion mirrors the usual continuity of linear functionals on topological vector spaces, but is tailored to the order‑theoretic context. The main representation theorem states that every continuous linear form φ can be uniquely expressed as a maxitive integral of the form

  φ(x) = ∫⁺ f·dμ

where μ is a regular maxitive measure on M (valued in L) and f is a “density” element of the semimodule (often a continuous function). The integral ∫⁺ is the idempotent analogue of the Lebesgue integral, defined via suprema of simple functions. The proof relies on the join‑density of L, the existence of a basis of compact elements, and the fact that Scott‑continuous maps are determined by their values on these compact generators.

Having established the representation, the authors turn to two classical theorems in measure theory and functional analysis, recasting them in the idempotent setting. First, an idempotent Radon–Nikodym theorem is proved: if a maxitive measure ν is absolutely continuous with respect to another maxitive measure μ (in the sense that ν(A) ≤ μ(A) for all A implies ν(A) = 0), then there exists a unique density f such that ν(A) = ∫⁺_A f·dμ for every measurable set A. The continuity of the underlying lattice ensures that the density can be taken to be a Scott‑continuous function, and the Z‑framework guarantees that the construction works uniformly for a wide class of semimodules.

Second, an idempotent Riesz representation theorem is obtained. For a space C_c(M, L) of compactly supported, Scott‑continuous functions from M to L, every continuous linear functional Λ on C_c can be represented uniquely by a regular maxitive measure μ via Λ(g) = ∫⁺ g·dμ. This mirrors the classical Riesz theorem for Radon measures, but the proof avoids any reliance on topology beyond the order‑theoretic compactness inherent in the domain‑theoretic setting.

The paper also discusses the necessity of the continuity hypothesis. When φ fails to be Scott‑continuous, the representation may break down; the authors exhibit examples where a linear form cannot be written as a maxitive integral, and they outline possible extensions using decompositions into supremum‑infimum pairs or by relaxing regularity conditions on μ.

Overall, the work unifies several strands of research—domain theory, idempotent analysis, and maxitive measure theory—by showing that continuity (in the sense of Scott) is the key property that enables representation theorems analogous to those in classical analysis. The Z‑framework serves as a powerful unifying language, allowing the authors to treat a broad spectrum of structures (including complete lattices, continuous posets, and even certain quantales) within a single theory. The results open pathways for further applications in optimization, tropical geometry, and theoretical computer science, where idempotent structures and maxitive measures frequently arise.