Authors: Thomas Stieve*, University of Arizona - Geography & Development
Topics: Cultural Geography, Political Geography, Social Theory
Keywords: Critical Discourse Analysis, Knowledge Production, Globalization, Wikipedia
Session Type: Paper
Start / End Time: 8:00 AM / 9:40 AM
Room: Balcony L, Marriott, River Tower Elevators, 4th Floor
Presentation File: No File Uploaded
Foucault situated knowledge within the framework of discourses, which are defined as “a set of meanings, metaphors, representations, images, stories, statements and so on that in some way together produce a particular version of events” (Burr, 1995, p. 48). Critical discourse analysis (CDA) focuses on how social and political dominance are replicated in discourses (Fairclough, 1992; Fairclough, 2001; Wodak 2001). The concepts of power, history and ideology are pivotal in this analysis. Moreover, the scope and scale of the discourse become vital (Fairclough 1993; van Dijk 2002). The emphasis in discursive investigations can be at a more local, situational level, where the production of the text is viewed within a specific social context. At the other end of the scale, one can study the mega-discourse, where discourses are more universally connected, examining phenomena such as globalization (Alvesson and Karreman 2000). Since its start in 2001, people from across the globe have produced knowledge in hundreds of languages online in Wikipedia. This methodological paper explores the most appropriate methods to investigate how knowledge as represented in this international encyclopedia is produced globally or remains local. Conventionally, big data, of which Wikipedia certainly is an example, involve algorithmic exploration such as latent Dirichlet allocation (LDA) while CDA traditionally examines topical online discourses (Brock, 2016). By focusing on the globalizing discourses of trade, mass media and the English language, I examine the differences in investigating globalized discourses by employing LDA and CDA on Wikipedian articles, specifically concentrating on text, images and links.