According to Gillian Rose, discourse is “groups of statements which structure the way a thing is thought, and the way we act on the basis of that thinking” (2001, pg. 136). In my opinion, discourse can be thought of as a conversation that’s being had about a certain issue or within a certain field. Discourses privilege certain viewpoints over others depending on the power relations – political discourse favours those who are politicians or those who work as political journalists over the layman’s opinions, for example. Discourse is not simply a sum of statements, it can also construct positions. Modern conceptions of gender identity are discursive constructions, for one. As Rose puts it, “Our sense of our self is made through the operation of discourse. So too are objects, relations, places, scenes: discourse produces the world as it understands it.” (pg. 137)
Rose is referencing Foucault’s understanding of discourse and his work on power/knowledge relations. In a nutshell, all knowledge is dependent on the presupposition of power relations which are simultaneously upheld by fields of knowledge.
In trying to understand the forces and implications of various discourses, Rose proposes 2 different methods for studying it. In the first form of discourse analysis, Rose is interested in the specific visual images and texts that constitute a discourse.
In her book chapter, she outlines a methodology for conducting the first form of discourse analysis. First, a researcher must identify the right sources from which to collect data. This might be a difficult task because the parameters must be decided by the researcher who must know when to stop and what is the right quantity to derive the most quality of data from. Next, the data must be assessed. According to Foucault’s work, a researcher must look at the information presented in the sources without preconceived notions and to look at them with fresh eyes. Once that is complete, a researcher should sort the information according to themes and codes in order to draw links between the data gathered.
Rose emphasises that it is important to study how a discourse produces the “effects of truth” (pg. 154) – how does discourse claim to represent the truth or establish certainty for the way things are? What is left out or rendered absent by the discourse? And how does the discourse deal with contradictory viewpoints? These are all questions a researcher must answer in order to conduct a full analysis of the data.
We can see Rose’s description of discourse analysis in use through Charlene Elliott’s 2001 study of the discourse of Starbucks and coffee. In her analysis, Elliott looks at the marketing and on-site copy produced by Starbucks and how they produce misleading stereotypes about foreignness and the “Third World” coffee producing nations. Through carefully studying the marketing materials of Starbucks, Elliott discovered that the brand had positioned beans from non-Western locales as “exotic” or “wild”, thus perpetuating Orientalist ideas about the Other. By positioning its premium blends as more foreign, Starbucks afforded Western consumers an opportunity to consume exotic locations in a neocolonialist fashion.
Elliott compared the diction of various blends of coffee and noted that house blends were described with words like “bright” or “balanced” while blends which were marked as being from non-Western places like Kenya or Sulawesi were ascribed terms like “exotic but approachable” and “intense” (pg. 377). She then compared the descriptions within the marketing copy to unexpected sources like postcolonial scholar Edward Said’s Orientalism to demonstrate exactly how Starbucks was exploiting the rhetoric of foreignness in order to market its products.