Kuroda's Translation for the $λΠ$-Calculus Modulo Theory and Dedukti
Thomas Traversié
TL;DR
This work addresses translating classical proofs into intuitionistic ones within the lambdaPi-calculus modulo theory, a framework that combines dependent types and user-defined rewrite rules in Dedukti. The authors extend Kuroda's double-negation translation to theories encoded in higher-order logic, formalize the embedding of classical into intuitionistic logic, and implement the translation in Construkti to produce translated Dedukti proofs. They demonstrate the approach by translating and benchmarking a substantial library of proofs (including propositional, first-order, and higher-order results) and report successful typechecking of all translated proofs. The results enable interoperability between classical proof systems and intuitionistic proof assistants, facilitating cross-system proof exchange and verification.
Abstract
Kuroda's translation embeds classical first-order logic into intuitionistic logic, through the insertion of double negations. Recently, Brown and Rizkallah extended this translation to higher-order logic. In this paper, we adapt it for theories encoded in higher-order logic in the lambdaPi-calculus modulo theory, a logical framework that extends lambda-calculus with dependent types and user-defined rewrite rules. We develop a tool that implements Kuroda's translation for proofs written in Dedukti, a proof language based on the lambdaPi-calculus modulo theory.
