Polarizing Double Negation Translations

Mélanie Boudard,Olivier Hermant
DOI: https://doi.org/10.48550/arXiv.1312.5420
2013-12-19
Logic in Computer Science
Abstract:Double-negation translations are used to encode and decode classical proofs in intuitionistic logic. We show that, in the cut-free fragment, we can simplify the translations and introduce fewer negations. To achieve this, we consider the polarization of the formul{\ae}{} and adapt those translation to the different connectives and quantifiers. We show that the embedding results still hold, using a customized version of the focused classical sequent calculus. We also prove the latter equivalent to more usual versions of the sequent calculus. This polarization process allows lighter embeddings, and sheds some light on the relationship between intuitionistic and classical connectives.
What problem does this paper attempt to address?