Substructural logics are formal reasoning systems that relax or drop some of the standard rules of classical logic, such as the ability to freely duplicate or discard information. They appear in computer science, linguistics, and the study of resource-sensitive reasoning. This paper extends one well-known substructural logic by adding two new logical operators that form a "Galois connection," meaning they interact in a precisely balanced, adjoint way, and that also satisfy additional "Frobenius" compatibility conditions. The Frobenius conditions, borrowed from algebra and category theory, essentially require that the two operators mesh coherently with the logic's existing connectives. The result is a new, richer logic whose behavior the authors then work to understand from multiple mathematical angles.
The first major contribution is showing that this new logic is "algebraizable," meaning there is a clean, provably equivalent translation between the logic and a class of algebraic structures the authors call Frobenius-adjoint residuated lattices. This is important because it lets logicians use algebraic tools to study the logic and vice versa. They also confirm that the new logic does not prove anything new about statements that don't mention the new operators, and they connect questions about whether the logic can always find finite counterexamples to purely algebraic properties of these structures.
The second and third contributions focus on the distributive case, where the underlying lattice structure behaves like sets under union and intersection. Here the authors lift a classical construction, due to Kalman, which normally relates a logic to a simpler "positive" version of itself, into this new Frobenius-adjoint setting. They establish a categorical equivalence, essentially a perfect structural correspondence, between two families of algebraic structures. As a payoff, they show that reasoning in the full logic, reasoning about the algebraic models, and reasoning about the simpler positive-cone models all yield exactly the same conclusions. This triple agreement holds not just for ordinary logical consequence but also for equational and quasi-equational reasoning, conservativity, and the existence of finite counterexamples, giving a unified and robust framework for this class of logics.