Home Institution

Harvard University

Publication Date

Fall 2023

Program Name

Portugal: Sustainability and Environmental Justice

Abstract

Expressions of environmental distress can take different forms, with different symptoms, causes, and treatments. Existing literature generally identifies three primary categories of environmental distress responses: environmental grief, solastalgia, and eco-anxiety. This paper explores these distinctions and identifies words typically associated with each category in Portuguese news articles using Latent Dirichlet Allocation (LDA). LDA is a Natural Language Processing (NLP) technique that groups words into topics and identifies connections between those words based on how often those words appear together in sequences. This paper uses a modified LDA algorithm called GuidedLDA to identify additional keywords within topics defined by a list of “seed” keywords. The coherence scores for two article sets analyzed with these categorizations were -0.87 and -1.05, indicating that these categorizations of environmental distress are clearly delineated in media. Trends identified in the LDA analysis matched findings from qualitative studies, such as the persistence of environmental grief as time-independent and the strong relationship between place attachment and solastalgia. These findings provide quantitative support for previous qualitative metrics and offer a framework for further examinations of environmental sentiments in news media.

Disciplines

Climate | Data Science | Environmental Studies | Journalism Studies | Mass Communication | Multicultural Psychology | Sociology of Culture | Spanish and Portuguese Language and Literature

Download

Share

Article Location

 
COinS