Abstract: We consider algorithmic properties of mathematical models used in computational linguistics to formalize and represent the semantics of natural language sentences. In particular, finite-order functionals play a crucial role in Montague intensional logic and formal semantics for natural languages. We compare several computable models for the spaces of finite-order functionals based on Ershov-Scott theory of domains and approximation spaces. Namely, we describe how complexity and representability of functional spaces depend from the choice of three basic domains: for entities, for truth values, and for states.

Cite: Stukachev A.
Computable models of Montague intensional logic
Fifth Workshop on Digitalization and Computable Models 02-06 Oct 2023