Type-Theoretical Semantics

Type-Theoretical Semantics

Formal Approaches to Meaning in Natural Language

Overview

CLLT conducts pioneering research in type-theoretical approaches to natural language semantics, bridging formal linguistics and computational implementation. Our work demonstrates how modern type theories—particularly Martin-Löf Type Theory and Coq—can provide rigorous foundations for semantic analysis.

This research stream combines deep theoretical work with practical computational implementations, creating systems that are both mathematically rigorous and computationally executable.

Key Publications

Formal Semantics in Modern Type Theories

Formal Semantics in Modern Type Theories

Chatzikyriakidis, S., & Luo, Z. (2020)

Comprehensive monograph on formal semantics using Modern Type Theories (MTT). Published by ISTE/Wiley Science, this work establishes MTT as a powerful framework for natural language semantics, covering compositionality, quantification, anaphora, and more.

Coq Implementation of FraCaS

Natural Language Inference in Coq

Bernardy, J.P., & Chatzikyriakidis, S. (2019)

Implementation of natural language inference using the Coq proof assistant. This work demonstrates wide-coverage symbolic NLI, providing executable formal semantics for reasoning tasks including the FraCaS test suite.

Collaboration with Leading Researchers

This research is conducted in collaboration with:

  • Zhaohui Luo (Royal Holloway, University of London) - Modern Type Theories
  • Jean-Philippe Bernardy (Chalmers/University of Gothenburg) - Coq implementations
  • Robin Cooper (University of Gothenburg) - Type Theory with Records

Applications

Natural Language Inference

Natural Language Inference

Symbolic NLI systems with provable correctness and explainable inference chains

Formal Verification

Formal Verification of Semantics Theories

Formal verifications of formal semantics theories