Academia.eduAcademia.edu

Basic Constructive Modality

Abstract

The benefits of the extended Curry-Howard correspondence relating the simply typed lambda-calculus to proofs of intuitionistic propositional logic and to appropriate classes of categories that model the calculus are widely known. In this paper we show an analogous correspondence between a simple constructive modal logic CK (with both necessity and possibility ♦ operators) and a lambda-calculus with modality constructors. Then we investigate classes of categorical models for this logic. Parallel work for constructive S4 (CS4) has appeared before in . The work on the basic system CK has appeared initially with co-authors Bellin and Ritter in the conference Methods for the Modalities [Bellin et al., 2001]. Since then the technical work has been improved by and taken to a different, higher-order categorical setting by Ritter and myself. Here we expound on the logical significance of the earlier work. liminary work on a dual system for intuitionistic and (S4) modal logic, called DIML for Dual and Intuitionistic Modal logic (joint with Neil Ghani and Eike Ritter ]) came to light. By then I was truly bitten by the bug: I wanted to see how far we can push the frontiers of the Curry-Howard correspondence for modal systems. But I wanted my Curry-Howard correspondence to be a categorical one, that is, I wanted "triangles" of maps relating logics, their type-theoretical formulations and their (equivalent) categorical semantics formulations. With the help of Natasha Alechina, Eike Ritter and Michael Mendler