Toward Intuitive Programming Languages

Abstract

Modern text-based computer programming languages use syntax, semantics, and APIs to represent actions a computer will perform. Historically, the design of these languages has largely escaped the critical gaze of peer review, relying more on expert opinions than robust scientific methodologies. In this paper, we pose a questionâ€"is it possible to create a programming language where the syntax, semantics, and API design is based upon rigorous data collection and the scientific method? We have undertaken a long-term project to develop a computer programming language, called Hop, where each language decision is based upon empirical metrics gathered from human studies. While such a design may not universally benefit all programmers, our hope is that such a procedure may make our community's language design decisions more objective and transparent.