Software and Stylistics

Context: Texts, and a very wide variety of texts, are often the medium through which software engineers articulate abstractions and express their thinking, both to each other and to the computational machine. It is therefore important that the Software Engineering community understands what is happening with these texts. Stylistics brings together linguistics, literature and cognitive science to understand how stylistic features and patterns in texts give rise to human interpretation and meaning. Stylistics and Software Engineering therefore share a common interest in the close relationship between code and meaning/effect. (Stylistics’ conception of code subsumes computing’s conception.) Because of the common interest in the features and patterns of texts, we hypothesise that the disciplines of Stylistics and Software Engineering have much that they can contribute to each other. Furthermore, no previous research project has explored the relationship between Stylistics and Software Engineering, or the contributions that each discipline might make to the other, in the way we explore them here. Thus, the cross-disciplinary project is, itself, a very novel exploration. Aim and Objectives: The primary aim of the proposed discipline hop is to establish strong, shared, cross-disciplinary intellectual and technical foundations between the Project Lead and the Project Co-lead as preparation for a future series of research projects that bring together Stylistics and Computing, particularly Software Engineering. The secondary aim is to demonstrate to the respective research communities (local, national and international) the extent to which cross-disciplinary research between Stylistics and Software Engineering is both achievable and worthwhile. The objectives for the project are to: Explore the contribution that Stylistics can make to Software Engineering (SE). The project will focus in three areas of SE: stories in requirements engineering; program comprehension in the Modern Code Review; and the use of large language models (LLMs) to 'translate' input texts into output texts for requirements and code. Explore the contribution that Software Engineering can make to Stylistics. The project will focus in two areas: the variety and nature of texts used in SE as a novel source of data for Stylistics, and computational thinking, i.e., how software engineers think and how they use texts in that thinking. Co-create one or more substantive, demonstrable examples of the integration of Stylistics and Software Engineering. Promote the integration of (aspects of) Stylistics and Software Engineering, locally, nationally, and internationally. The core of the project is structured around four 4-month mini-hops, with two mini-hops per investigator, over a 24 month period. The project will comprise of formal study of the respective discipline, engagement and facilitation, co-creation of interdisciplinary outputs, and promotion of the bringing together of Stylistics and Software Engineering. The direct impacts of the project are expected to be the development of cross-disciplinary knowledge across the respective research disciplines of SE (e.g., designing studies of SE based on principles of Stylistics) and Stylistics (e.g., studies of GitHub projects as corpora). Indirect, longer-term, routes-to-impact are expected in SE practice (e.g., cognitive guidelines for software engineers on how to write and read requirements specifications, and cognitive guidelines on how to review source code) and in Stylistics (e.g., guidance on accurate translations of texts, and guidance on forensic analysis of texts for determining authorship, including that of human vs machine).