Critical and descriptive goals in discourse analysis

Abstract

I view social institutions as containing diverse 'ideological-discursive formations' (IDFs) associated with different groups within the institution. There is usually one IDF which is clearly dominant. Each IDF is a sort of 'speech community' with its own discourse norms but also, embedded within and symbolized by the latter, its own 'ideological norms'. Institutional subjects are constructed, in accordance with the norms of an IDF, in subject positions whose ideological underpinnings they may be unaware of. A characteristic of a dominant IDF is the capacity to 'naturalize' ideologies, i.e. to win acceptance for them as non-ideological 'common sense'. It is argued that the orderliness of interactions depends in part upon such naturalized ideologies. To 'denaturalized' them is the objective of a discourse analysis which adopts 'critical' goals. I suggest that denaturalization involves showing how social structures determine properties of discourse, and how discourse in turn determines social structures. This requires a 'global' (macro/micro) explanatory framework which contrasts with the non-explanatory or only 'locally' explanatory frameworks of 'descriptive' work in discourse analysis. I include a critique of features of such work which follow from its limited explanatory goals (its concept of 'background knowledge', 'speaker-goal' explanatory models, and its neglect of power), and discuss the social conditions under which critical discourse analysis might be an effective practice of intervention, and a significant element in mother tongue education. © 1985.