Picture from a cool article by T. Mills Kelly at The Conversation
In the chapter "Linguistic Analyses of Language Policies", Ruth Wodak suggests that a discourse analysis method can be helpful in studying language policies (2006: 170-193), so I decided I would try to implement this in my research. To prepare, I read How To Do Critical Discourse Analysis by David Machin and Andrea Mayr (2012).
One thing that has become clear from reading the book is that Wodak's approach is much more extensive and thorough than that described by Machin and Mayr. They see her work as an extension of an approach that is already a bit wider than the "common" critical discourse analysis (CDA) method, namely Van Dijk's socio-cognitive approach. In the "common" method, as described by Machin and Mayr, a few texts are picked out that are deemed representative of a particular kind of discourse by the analyst. He or she then analyses what semiotic choices are made that conceal the underlying ideology of the text. The purpose of CDA is to bring out and challenge hidden power relations, which they believe are created and reinforced through language and through specific discourses.
In my post on "discourse dissemination across social boundaries" I talked about some of my hesitations with this approach, such as that it did not take into account who produced the text for which kinds of readers, and how those readers react to it. The conclusion of Machin and Mayr's book actually goes into these kinds of criticisms; that it does not consider the background and context of the production and consumption of the texts, on the one hand, and that it is too selective and subjective on the other hand.
In the discourse-historical approach described by Wodak, these issues are resolved to an extent. Background reasearch and ethnography are used to provide the actual context and history of certain discourses, to understand how they come about and what function they serve. Because of the large amount of data that are gathered, the selection problem of CDA is less acute of a problem. CDA becomes less of a focus on its own and more of a tool used to show how the discourses work in practice. It is also suggested that quantitative analysis (e.g. Corpus-based) be integrated, which would make the interpretation less subjective.
As a whole the approach probably makes for a PhD-size project per proper implementation. The next step for me is to think about how I could realistically integrate this into my own project. I'm now interested in ethnography and corpus linguistics though, and I'll be reading up about those. Also for next week I will try to apply my new CDA knowledge onto some of the articles I posted on here before.
