Multirole Logic

Author: Hongwei Xi and Hanwen Wu

© 2016-2018 Hongwei Xi and Hanwen Wu, All Rights Reserved.

We identify multirole logic as a new form of logic in which conjunction/disjunction is interpreted as an ultrafilter on the power set of some underlying set (of roles) and the notion of negation is generalized to endomorphisms on this underlying set. We formalize both multirole logic (MRL) and linear multirole logic (LMRL) as natural generalizations of classical logic (CL) and classical linear logic (CLL), respectively, and also present a filter-based interpretation for intuitionism in multirole logic. Among various meta-properties established for MRL and LMRL, we obtain one named multiparty cut-elimination stating that every cut involving one or more sequents (as a generalization of a binary cut involving exactly two sequents) can be eliminated, thus extending the celebrated result of cut-elimination by Gentzen.

Papers

Talks

The following talk was given by Hanwen Wu at IBM PL Day 2017 hosted by IBM T.J. Watson Research Center. (Dec 4. 2017)

The following talk was given by Hongwei Xi and Hanwen Wu at LC 2017. (Aug 18, 2017)

A version with speaker’s notes can be found here on Scribd.

The following talk was given by Hanwen Wu at NEPLS 2016 hosted by UMass Amherst. (May 31, 2016)

The following talk was given by Hanwen Wu at NEPLS 2015 hosted by Tufts University. (Nov 10, 2015)

It’s code can be found here which also includes a simple encoding of ConcurrentML.

Full Proofs

These proofs are formulated and checked in an older version of ATS and z3.1 We recommend that you verify one proof at a time and comment out others to make it possible to verify within resource limit.

Implementations

Our work has enabled several implementations. The session API formulation is done in Applied Type System. After type checking, it will be compiled into a target language. Currently we experimented several such targets, including C, JavaScript, and Elixir/Erlang.

Other Resources


  1. It may broke if the version is seriously outdated.