Linear logic

Linear logic
Name	Linear logic
Introduced	1987
Creator	Jean-Yves Girard
Field	Proof theory, Category theory, Computer science

Linear logic is a substructural logic introduced in 1987 by Jean-Yves Girard as a refinement of classical and intuitionistic logics that controls resource usage. It reinterprets structural rules of Gentzen-style proof systems, distinguishing between reusable and consumable propositions and establishing correspondences with category theory, automata theory, and lambda calculus. The framework rapidly influenced work in proof theory, type theory, programming language theory, and concurrency theory.

History

Girard presented the system at a time of active exchange among researchers in proof theory, lambda calculus, and category theory; contemporaneous institutions included CNRS, École Normale Supérieure, and conferences like LICS. Early adoption intersected with work by researchers at Princeton University, University of Paris, and MIT, and spurred follow-up by figures associated with Gödel Prize-level contributions. Developments connected to earlier strands such as Gentzen, Sequent calculus, and the study of structural rules traced through collaborations involving scholars tied to Stanford University, University of California, Berkeley, and INRIA.

Syntax and sequent calculus

The formal syntax uses multiplicative and additive connectives, exponential modalities, and units: multiplicatives (tensor ⊗, par ⅋) and additives (with &, plus ⊕) alongside exponentials (!, ?), plus units (1, ⟂, ⊤, 0). The sequent calculus developed in the late 1980s and 1990s refines Gentzen-style sequents, introducing controlled use of weakening and contraction via exponentials; influential presentations appeared in textbooks associated with Cambridge University Press and lecture notes from ENS Lyon and University of Oxford. Proof systems separate rules for left and right introduction for each connective, and use structural contexts that reflect resource sensitivity studied in seminars at IHÉS and workshops at CWI.

Semantics and models

Semantic accounts include phase semantics, coherent spaces, and categorical models such as *-autonomous categories and linearly distributive categories. Phase semantics were introduced in Girard's foundational work and elaborated by researchers at University of Cambridge and École Polytechnique. Coherent spaces connected the logic to denotational models developed by groups at University of Paris-Sud and INRIA Rocquencourt. Categorical semantics use structures related to Monoidal category, adjunctions, and *-autonomy; important contributors include researchers from University of Edinburgh and Rutgers University. Game semantics and geometry of interaction provide operational models explored at CNRS, University of Toronto, and in collaborations involving Carnegie Mellon University.

Proof theory and cut-elimination

Cut-elimination for linear logic generalizes Gentzen's cut-elimination theorem and underlies consistency and normalization results; proofs were advanced in seminars at IHÉS and formalized in monographs from Springer. The proof-theoretic study examines focalization, normalization, and proof nets: proof nets offer a parallel syntax whose correctness criteria were developed by teams at Université Paris 7 and University of Cambridge. Geometry of Interaction, introduced by Girard, gives a dynamic interpretation of cut-elimination with implementations influenced by groups at École Normale Supérieure and University of Pennsylvania.

Variants and extensions

Variants include Intuitionistic linear systems, Classical linear systems, Differential linear logic, and Non-commutative and ordered variants. Differential linear logic was developed by researchers affiliated with Université Paris-Sud and collaborators in the European Research Council network. Extensions incorporate modalities and quantitative refinements studied in projects at MIT, Harvard University, and University of Toronto. Connections to modal logic frameworks prompted cross-disciplinary work with groups at University of Cambridge and Oxford University Press-sourced texts.

Applications

Linear logic has been applied to type systems, concurrency, and resource-sensitive computation: linear type systems informed language design in projects at Bell Labs, Microsoft Research, and academic labs at Princeton University. Linear logic underlies session types in process calculi explored by researchers at INRIA and Carnegie Mellon University, and models for compiler optimization and memory management by teams at Google and IBM Research. It influenced formal verification tools developed in collaboration with National Institute of Standards and Technology and European research labs, and informed semantics in quantum computation research pursued at Perimeter Institute and University of Waterloo.

Examples and fragments

Key fragments include Multiplicative Linear Logic (MLL), Multiplicative-Additive Linear Logic (MALL), Intuitionistic Linear Logic (ILL), and Affine and Relevant fragments. MLL and MALL were studied extensively at seminars at University of Cambridge and University of Oxford; ILL connects to work on the Curry–Howard correspondence investigated by groups at Princeton University and University of Edinburgh. Concrete examples include encoding of stateful computation and reversible computation abstractions explored in collaborations involving MIT and École Polytechnique.

Category:Logic