Re-storying Environmental Policy: An Ecological AI-Assisted Approach

In recent years, growing awareness of the cultural roots of the socioecological crisis has led to a major scholarly interest in how language mediates socio-ecological relations. While ecolinguistic research has extensively documented how linguistic patterns promote destructive or beneficial narratives, experimental evidence on how exposure to different discursive configurations shapes readers’ environmental perception and attitudes remains limited. This study aims to mitigate this gap by investigating how semantic and syntactic variations in policy texts may influence readers’ environmental perceptions. A questionnaire was administered in summer 2025 to a sample of Chinese environmental science graduate and doctoral students, who evaluated three versions of an environmental policy document, re-storied using an AI-assisted protocol. The questionnaire was based on the New Ecological Paradigm (NEP) scale, adjusted to serve as an evaluative tool for assessing the perceived texts’ environmental positioning. Results show statistically significant differences across assessments provided for the three documents, highlighting a correlation between linguistic variations and readers’ interpretation of texts. From a methodological standpoint, this study provides empirical confirmation of NEP’s validity as a tool for ecolinguistics assessment of texts relevant for environmental decision-making. Furthermore, it demonstrates the methodological potential of generative AI as an ecolinguistics-informed re-storying tool.