Irene Elmerot antogs som doktorand inom forskarskolan Språk och makt 2018. Hennes korpusstödda avhandling Decoding Discourse: A corpus linguistic study of evaluative adjectives and group nouns in Czech print news media (1989–2018) beskriver hur olika grupper av människor har representerats i tjeckisk nyhetspress under tre årtionden efter Sammetsrevolutionen, med hjälp av adjektiv klassificerade enligt Subjectivity Lexicon for Czech (Veselovska & Bojar 2013). Sammanläggningsavhandlingen är en intersektionell studie med grupper av nationaliteter och etniciteter, kön samt yrkesgrupper. Det förekommer även variabler för religiositet och bruttonationalinkomst i datasetet.

Disputationen kommer att vara på engelska.

Mer information kommer.

Läs mer om Irenes avhandlingsprojekt

Avhandlingen i fulltext i Diva

----

Abstract [en]

This compilation thesis takes a top-down perspective on the representation of different groups of people in Czech news press over three decades. The starting point is that human equality is a global prerequisite for a democratic world, according to the United Nations Sustainable Development Goals. The research questions for the thesis concern how positively or negatively different groups of people are represented, and how often the different groups appear compared to each other. The thesis contributes results based on a language other than English, which represents a valuable contribution to the field.

Theoretically, focus is on the premise that language is a tool for gaining and maintaining power, and a way of expressing power relations (Reisigl & Wodak 2016; Fairclough 2015). An important theoretical focus is the phenomenon of linguistic othering (Fidler 2016), which here means letting a group of people stand out from a certain news description by emphasising some of their characteristics. They then form what is also called an outgroup, as opposed to the ingroup that the writer is assumed to be part of (van Dijk 1987). The findings of this thesis provide insights into how news media can influence our perceptions of, for example, different nationalities or professions, linked to their socio-economic status, and by extension how these perceptions can influence our attitudes and behaviours towards these groups.

Methodologically, the thesis uses corpus-based discourse analysis. Empirically, the research is based on the Czech National Corpus (www.korpus.cz). From this corpus, 32 million observations are extracted of when positive and negative adjectives, classified according to a subjectivity lexicon, appear in the news press together with nouns for different kinds of groups of people, such as gendered words like “woman” or “man”, occupations like “maid” or “miner” and nationalities like “Somali” or “Dane”. When adjectives are closer to nouns, or even next to them, they are given more weight than when they are more distant (Cvrček 2014). With such large amounts of data, a top-down or bird’s eye view is the most reasonable, but some detailed analyses are also included.

Study I focuses on the representation of nationalities and countries, classified by the World Bank into groups according to their gross national income, and their co-occurrence with the positive and negative adjectives. Results: the nationalities in the different income groups are represented in a descending order; the higher the national income, the more positive the representation. Furthermore, discourses related to the so-called war on terror, as well as the security of different nations, emerge as a result of the analysis.

Study II focuses on two groups, a focus group of Arabs and Muslims and a reference group of the other nationalities and countries. The focus group is a very heterogeneous group of people and countries that is often portrayed in the context of conflict (Baker et al., 2013, pp. 2 and 32). Results: Arabs and Muslims are consistently represented as an out-group, which over time affects how the people who read these news media view them.

Study III contains two sub-studies, based on an intersectional analysis of modern Czech news reporting; in one sub-study the analysis focuses on professional roles, and in the other on different nouns for women and men. Results: Those with lower socio-economic status and fewer supervisory roles in their work are less likely to appear in news coverage, but when they do appear, it is not always with more negative representations. Regarding gender, men are more often portrayed than women, and women are more often represented by evaluative adjectives than men. In addition, women’s positive representations are based to a greater extent on their appearance and feelings, while men’s representations are based on their importance and competence.

Overall, the results confirm quantitatively, with an empirical material covering almost the entire print news reporting in the Czech Republic since democratisation, that hypotheses that have been theoretically proposed, as well as confirmed, for other countries, turn out to be true for Czech news reporting. There are systematic differences in the way that some groups of people are significantly more often represented in the media than others, and that some groups are systematically represented more favourably than others. It also shows that these imbalances are clearly linked to factors such as nationality, occupation and gender.