ABSTRACT
Practice theorists such as Bakhtin (1981), Bourdieu (1977) and Volosinov (1973) have all conceptualized language as constituting and embodying ideology.2 In this view (Philips 1992: 378), language is `itself material' rather than independent and simply symptomatic of material realities. Scholarship on the language of institutional settings has given this theoretical view empirical substance: researchers have investigated the role of discursive practices in constructing and constituting power relations among professionals and clients, in particular, the interactional mechanisms by which certain ideological or interpretive `frames' dominate institutional interactions, while others are suppressed (Philips 1992). Todd (1989) and Fisher (1991), for example, document how doctors' medical and technical concerns prevail in interactions with patients, even when patients articulate their problems in social and/or biographical terms. In her comparison of a doctor±patient interaction and a nurse-practioner±patient interaction, Fisher (1991: 162) isolates aspects of interactional structure related to such discursive control: the doctor, much more than the nurse-practitioner, asked questions that `both allow[ed] a very limited exchange of information and le[ft] the way open for his [the doctor's] own assumptions to structure subsequent exchanges'. By contrast, the nurse-practitioner used open-ended, probing questions which maximized the patient's own `voice' and interpretation of medical problems. In the context of legal settings, Walker (1987: 79) also comments on the interactional control exerted by questioners. As part of her investigation of linguistic manipulation in legal settings, she interviewed witnesses who reported `a feeling of frustration at being denied the right to tell their own stories their own way'. While the stories told in court were the witnesses' own, lawyers and judges had the socially-sanctioned prerogative to `present, characterize, limit and otherwise direct the ¯ow of testimony' such that the frames structuring the stories were often not the witnesses' own. In Fisher's (1991: 162) terms, `both the questions and the silences ± the questions not asked ± do ideological work.' Not only was Fisher's doctor±patient interaction structured by the doctor's assumptions (due to questions that
allowed a limited exchange of information), but implicit in these assumptions were views about the centrality of the nuclear family to this mother's sense of well-being or ill-health. According to Fisher, the doctor's questions functioned to reinscribe the hegemonic discourse that `justif[ies] the traditional nuclear family which has at its center a mother' (Fisher 1991: 162, emphasis in original).
