Language Matters in British Newspapers: A Participatory Analysis of the Autism UK Press Corpus

Abstract

Background: Language around autism plays a crucial role in shaping public attitudes toward autistic people. The use of identity-first versus person-first language and impersonal references to autism can affect how autistic people are perceived. These factors should impact the representation of autistic people in newspapers, where negative and stereotypical representations are often perpetuated. Method: We asked five autistic people to judge the sentiment toward autism and autistic people in 1000 quotes from British newspapers (2011–2020). The coders, who did not know the newspaper title and time of publication, made their judgments based on two dimensions, warmth and competence, from the Stereotype Content Model (SCM). We examined the overall judgments of warmth and competence and considered variations in language context and terminology, such as the use of impersonal references to autism or identity-first and person-first language. We also examined potential differences between broadsheets and tabloids, left- and right-leaning newspapers, and changes over time. Results: The majority of quotes from British newspapers fell under a low warmth and low competence area within SCM. Furthermore, impersonal references to autism tended to be rated lower in warmth and competence than references linking autism to an individual, whereas identity-first language was judged higher in warmth and competence than person-first language. Quotes from broadsheets were assigned similar warmth and slightly higher competence than quotes from tabloids. However, left-leaning and right-leaning papers did not differ regarding warmth and competence, and there were inconsistent changes over time. Conclusion: Our study confirms that the portrayal of autistic people in British newspapers tends to be negative. According to autistic raters, associating autism with a person and using identity-first language are linked to more positive representations. Although we found subtle variations in sentiment related to reporting style, our study shows little progress over time toward more positive portrayals.