Mind the Gap: General-Purpose Programming Languages Impede Scientific Model Development and Communication (PROPL 2026)

Computational models increasingly serve as the primary scientific artefacts in climate and earth-system research, yet the general-purpose languages used to build them often fail to express key domain semantics. We argue that this limits scientists' ability to represent and verify domain-specific concepts and invariants even as the community pushes toward more open, modular, and interoperable modeling ecosystems. We therefore advocate programming-language designs for computational science that make such domain meaning explicit and compositional, for example through intersection types or graded type systems.

Talk by Dominic Orchard (University of Cambridge and University of Kent) at the PROPL 2026 workshop at PLDI 2026 in Boulder, Colorado on 15th June 2026. Mirror of https://www.youtube.com/watch?v=8hc93jmlgvY.