Social scientists are not interested in identity in the sense of an individual’s unique name and address. They are interested in identity in the sense of the category that an individual belongs to (or is made to belong to). All languages have explicit names which allocate people to a category of person (e.g., madre in Spanish, umm in Arabic, or mother in English), an occupation (ingeniero, mohandass, enginee, respectively), or a position in society that has loose definitional criteria (gamberro, ’ozbagui, hooligan).
Much quantitative social science was (and is) devoted to “cover-sheet” identities like those, or still more general ones like gender, race, and class, and how they correlate with such variables as income, health, education, and crime. Social scientists who do qualitative research, however, take another tack. They believe that what is interesting about identities is how they are constituted – how society invents and perpetuates them. The way that society categorizes people, the laws it draws up, the visual images it promotes, the jokes it allows – all these are discourses of identity.
These discourses can be mapped in a descriptive way, to make a list of the features that society ascribes to any given identity category (one could, e.g., count the number of times the word “teenager” co-occurs with references to education, sex, television, and so on in the newspapers). Because that gives a rather static snapshot of the identity category, many social scientists are turning instead to discourse as an active social practice – i.e., to seeing discourse not as static, but as doing some business rhetorically or interactionally. The emphasis is not on the list of features of the category, but on what it means to bring up the category at that place and that time, and in those terms. Most of that sort of research falls under the heading of discourse analysis, which is a very varied set of analytic practices. Its five core features, applied to the study of identity, are these: (1) identities are to be understood not as essential and unchanging, but as subject to active construction, and liable to be imposed and resisted; (2) they appear in talk or text that is naturally found (in the sense of not invented or imagined by the researcher); (3) the identity words are to be understood in their co-text at least, and their more distant context if doing so can be defended; (4) the analyst is to be sensitive to the words’ non-literal meaning or force; (5) the analyst is to reveal the consequences achieved by the identities conjured by the words’ use – as enjoyed by those responsible for the words, and suffered by their addressees, or the world at large.
There are many ways in which those five criteria can be realized in the study of identity. I shall set them out in terms of their primary data; i.e., whether the discourse analyst looks for identities in people’s own words in interviews, in the context of a locally researched site, in the wider political context, or in the organization of interaction.
The Interview As Research Method
Interviews are still probably the method of choice for most qualitative researchers in the social sciences, on the proposition that they give respondents the freedom to express themselves, while at the same time allowing the researcher to probe specific questions that interest them. How one sets up the interview, and how one analyzes the resulting conversation, are a source of some debate in the literature. At one end are those who prefer the structured set of questions, and who treat the interviewee straightforwardly as an informant: thus a folklorist or oral historian might ask (say) community elders about their memories of their experiences as new immigrants, or about the customs of their particular sub-culture, and so on. The result is an enriched sense of identity issues as they are experienced by the people themselves, albeit filtered through memory and the demands of a retrospective interview.
At the other end are those researchers who look to the interview to provide evidence for further analytic work. The respondent’s story might be subjected to a thematic analysis, a narrative analysis, an interpretive phenomenological analysis, or a free association method. These analyses differ among themselves, but they all look to find something “between the lines” of the respondent’s account. A thematic analyst will look over a long account, or perhaps a set of accounts from different respondents, to find common themes of meaning – ways in which the respondent is making links between various parts of their identity. A narrative analyst will see how the respondent set her or his story out as a coherent narrative, using devices familiar from fiction to “plot” their life and people it with characters and events. The interpretive phenomenological analyst will use the interviewee’s words to try to get “inside their head” and understand their phenomenological experience of themselves. A free association narrative method will approach the respondent’s words from a psychoanalytic perspective, trying to understand the inner forces that produce the person they express themselves to be.
Research In Local Context
The ethnographically minded discourse researcher will be at pains to locate people’s words in their context and culture. Interviews may form part of the researcher’s toolkit, but principally as a way of informing the researcher of local meanings and codes; the analysis will be done on recordings of language as it is actually used, and identities as they come into play. Researchers in the ethnography of communication tradition, e.g., will look to see how speakers of different cultural identities signal those differences to each other, and how that affects (not always benignly) the success of their interaction.
Researchers in an ethnomethodological tradition will eschew interviews, and indeed any on-site material that is not brought actively into play by the speakers and participants themselves. The focus is on the ways that people publicly perform their reasonings about the world, such that they can prosecute their business with others: their reasoning is perforce available to all, without further interpretation or recourse to information not in the scene. So, e.g., in analyzing the way that a school student is judged to have a learning disability, Hugh Mehan (1996) analyzes not just the expert talk of the psychologist and the comparatively powerless talk of the parent, but also the props and furniture of the scene: the psychologists’ notes and statistical tables, the boardroom and table, the seating arrangements, and so on.
Research In A Political Context
Some discourse analysts prefer to approach their data from a given perspective on power and ideology in society, arguing that one must understand the social conditions of the production of a text before one can start reasonably to analyze it. The umbrella term “critical discourse analysis” shelters a broad family of analysts, but all are concerned to unpack the operation of power and ideology in discourse. Identities, their argument runs, are not always innocent, nor always under the control of the person identified; they can be oppressive, and the burden can be resisted.
Critical discourse analysts will not rely exclusively on one method or source of data. They may use research interviews with respondents, but their more common source of data on identities will be such “found” sources as newspaper and media accounts, political speeches, news interviews, and so on. The guiding principle will be to use linguistic and rhetorical scholarship to identify the metaphors, allusions, argumentative strategies, terminological choices, and pragmatic implications that converge to promote a certain vision of a group of people as falling under this or that socially significant identity. For example, a controversial identity like “asylum seeker” may be analyzed for how certain groups in society construct it to have – and other groups resist it having – politically and culturally negative implications. The work of the critical discourse analyst is to find subtle expressions of racism and power, and to unravel their workings.
Research In An Interactional Context
Among those methods of analyzing discourses of identity that look first to what happens in interaction, the one that pays closest, and most exclusive, attention to the precise sequence of talk is conversation analysis (CA). Its genesis was in the dissatisfaction of some sociologists in the late 1960s with the then dominant quantitative methodologies of their discipline, which were silent about how people actively realized the social world, in real time. In the 40 years since the pioneering work of the group around Harvey Sacks (whose lectures were published posthumously as Sacks 1992), CA has attracted a good deal of attention within sociology and outside it, and has developed into a multidisciplinary enterprise.
CA’s account of identities is to see them as being brought into active service in the particular moment of the interaction, for particular interactional effects. For a person to “have an identity” – whether he or she is the person speaking, being spoken to, or being spoken about – is to be cast (by a “membership categorization device,” which may be a term like “family” or the name of a profession, and so on) into a category with associated characteristics or features (the sort of thing you’d expect from any member of that category; their actions, beliefs, feelings, obligations, etc.). The force of “having an identity” is its consequentiality in the interaction – what it allows, prompts, or discourages participants to do or from doing next. For example, Dennis Day’s work (1998) on how people may allocate fellow-speakers to an ethnic group uncovers how mere hinting is enough to imply that someone is “not one of us,” and oblige them to take evasive action.
For CA, there is no need to go to the sort of abstract level favored by other kinds of discourse analysis, in their quest for the significance of identity talk: the speaker or writer’s use of (or hint at) an identity category is locally effective. If you call someone an asylum seeker (or hint that she or he is one) then you are doing it for local consumption, and the consequences will be interactionally visible. And this is true for mundane categories (like, say, “daughter”) as much as it is for more politically charged ones. In common with other discourse approaches, CA sees identities as constructed by language; its special emphasis is on language in use, at particular times and places.
- Antaki, C., and Widdicombe, S. (1998). Identities in talk. London: Sage.
- Benwell, B. M., & Stokoe, E. (2006). Discourse and identity. Edinburgh: Edinburgh University Press.
- Bamberg, M., & McCabe, A. (2000). Narrative identity. Special issue of Narrative Inquiry, 10(1).
- Day, D. (1998). Being ascribed, and resisting, membership of an ethnic group. In C. Antaki & S. Widdicombe (eds.), Identities in talk. London: Sage, pp. 151–170.
- Edwards, D. (1991). Categories are for talking: On the cognitive and discursive bases of categorization. Theory and Psychology, 1, 515 – 542.
- Edwards, D. (1998). The relevant thing about her: Social identity categories in use. In C. Antaki & S. Widdicombe (eds.), Identities in talk. London: Sage, pp. 15 – 33.
- Fairclough, N. (1995). Critical discourse analysis. London: Longman.
- Fitch, K. (1998). Speaking relationally: Culture, communication and interpersonal connections. New York and London: Guilford.
- Hollway, W., & Jefferson, T. (2000). Doing qualitative research differently: Free association, narrative and the interview method. London: Sage.
- Mehan, H. (1996). The construction of an LD student: A case study in the politics of representation. In M. Silverstein & G. Urban (eds.), Natural histories of discourses. Chicago: University of Chicago Press, pp. 253–276. Repr. in M.
- Wetherell, S. Taylor, and S. Yates (eds.) (2001). Discourse theory and practice. London: Sage, pp. 345 – 363.
- Sacks, H. (1992). Lectures on conversation, vols. 1 and 2. Oxford: Blackwell.
- Smith, J. A. (1999). Identity development during the transition to motherhood: An interpretative phenomenological analysis. Journal of Reproductive and Infant Psychology, 17, 281– 300.
- Tracy, K. (2002). Everyday talk: Building and reflecting identities. New York: Guilford.