Modal logic is a branch of formal logic that extends ordinary true/false reasoning by adding operators that express concepts like "necessarily" and "possibly." Traditionally, one way to give these operators precise mathematical meaning is through topology, where the "interior" of a set (roughly, its solid core, excluding boundary points) is used to model the necessity operator. This paper introduces a new variant called affine modal propositional logic, which modifies that classical setup by replacing the interior operator with a more flexible tool borrowed from analysis called the lower semi-continuous envelope operator. Instead of working with sets that are simply true or false, this framework works with functions that can take a range of values, which gives the logic a richer, more graduated structure suited to contexts where truth comes in degrees or where geometric and economic notions of convexity are relevant.
The key technical contribution is showing that this new logical system is both complete and "affinely compact." Completeness means that any statement which is true in all mathematical models of the logic can actually be proved using the logical rules alone, which is a fundamental sanity check for any formal system. Affine compactness is an analogue of a classical result saying that if every finite portion of a set of statements is logically consistent, then the whole collection is consistent too. These results together confirm that the logic is well-behaved and that the semantics faithfully captures what the formal proof system can express.
The motivation for this work comes from areas like convex analysis, optimization, and mathematical economics, where functions and their lower semi-continuous approximations play a central role. By building a logic directly around these objects, the paper offers a formal language for reasoning about properties that arise naturally in those fields. The affine modifier in the name signals that the logic respects linear and convex structure, distinguishing it from standard modal logics and opening potential applications wherever reasoning about convex or economic optimization problems requires a rigorous logical foundation.