Can GPT-4 learn to analyse moves in research article abstracts?

TL;DR

This study explores GPT-4's ability to automate the identification of rhetorical moves in research article abstracts, aiming to improve reliability and efficiency in discourse analysis.

Contribution

It demonstrates that GPT-4, with carefully designed prompts, can effectively identify moves in abstracts, reducing subjectivity and the need for multiple human coders.

Findings

8-shot prompts outperform 2-shot prompts in move recognition

Including variability examples enhances GPT-4's accuracy

GPT-4 shows potential for automating discourse annotation

Abstract

One of the most powerful and enduring ideas in written discourse analysis is that genres can be described in terms of the moves which structure a writer's purpose. Considerable research has sought to identify these distinct communicative acts, but analyses have been beset by problems of subjectivity, reliability and the time-consuming need for multiple coders to confirm analyses. In this paper we employ the affordances of GPT-4 to automate the annotation process by using natural language prompts. Focusing on abstracts from articles in four applied linguistics journals, we devise prompts which enable the model to identify moves effectively. The annotated outputs of these prompts were evaluated by two assessors with a third addressing disagreements. The results show that an 8-shot prompt was more effective than one using two, confirming that the inclusion of examples illustrating areas of variability can enhance GPT-4's ability to recognize multiple moves in a single sentence and reduce bias related to textual position. We suggest that GPT-4 offers considerable potential in automating this annotation process, when human actors with domain specific linguistic expertise inform the prompting process.