Conversation Analysis

Conversation analysis (henceforth, CA) is the study of social action as achieved through the medium of talk in interaction. Its genesis was in the dissatisfaction of some sociologists in the late 1960s with the dominant quantitative methodologies of their discipline, which were silent about how people actively realised the social world, in real time. In the forty 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 (for an account of the history of Conversation Analysis, see Heritage 1984; for a more recent overview of its methods and style, see Hutchby and Wooffitt, 1998; and for an account of its relation to other modes of discourse analysis, see Wooffitt, 2005).  

CA abides by the four generic DA criteria of looking for natural data, setting it in its cotext, watching for its non-literal meaning, and identifying the social actions performed. Perhaps its most obvious departure from this basic platform is its insistence on seeing social actions done though the very close organisation, as well as the content, of talk. In describing those actions, CA – again unlike generic DA – wants to stay as close as possible to the speakers’ own understandings of the actions without imposing interpretation from above or speculation about motives from below. Its ‘added value’ is teasing out the what and the how, while shying away from the why, and leaving off anything not made ‘live’ by the participants in the scene.

The currency that CA trades in might be structures on a chronologically minute scale (for example, the binding relation between speakers’ adjacent utterances, and the injunction to keep their separation brief) or extensive (the overall shape of a story delivered over many turns), but they are all normative. That is to say, speakers are expected to follow them, or risk (or invite) listeners to draw implications when they do not. We can see an example of such a normative structure in the simple example below, where the second utterance meets the expectation of a prompt acceptance of the first:

To show how strong the normative expectation is that the response be positive and prompt, consider this variant. Here the speaker’s non-normative silence in line 3 invites the listener to draw a significant implication.

Note that it is A who is responsible for both turns – so why does s/he answer his or her own question, and answer it with a negative? Because B has done the unexpected thing of not answering, and thus allowed the implication that the answer is ‘no’. A then makes plain that this has been understood. The interaction can proceed, with both parties now having disposed of the possibility that A visit B’s office on Monday, without A having had explicitly to say no. The ‘action’ has been achieved by exploiting of the regularities of talk.

CA has been applied productively to a variety of institutional activities otherwise accessible only in retrospect (by interviews with participants) or in simulation, or through comparatively coarse contemporary observation. For example,  CA has been used in research on how talk in interaction achieves business meetings (Boden, 1994), educational testing (Maynard and Marlaire, 1992) and survey interviewing (HoutkoopSteenstra, 2000), to take a few notable examples. 

What can CA reveal about such working interactions? Peräkylä and Vehviläinen (2003) put it neatly. Members of a trade or profession (they were talking about psychotherapists, but it’s true of anyone who routinely has dealings with clients) may have “stocks of interactional knowledge” -fairly clear ideas of what they do with the people they work with.  CA can check these accounts, correct them, or go beyond them.   In going beyond lay accounts, CA can  discover things about the interaction that the practitioners didn’t suspect, or which have effects or functions which don’t figure in (or indeed may be counter to) the official aims of the encounter. 

As an example of CA’s illumination of professional practice, consider Maynard’s work on clinicians’ delivery of a diagnosis. He inductively finds a pattern in which the clinician prefaces the actual diagnosis (you have X) by evidence (from test results, and so on). The typical sequence is like this, in which a doctor in a Developmental Difficulties clinic is talking to a mother about her five year-old son:

The doctor is describing evidence: the boy seems to function comfortably at the level of a three and a half year old. She is not (yet) giving a diagnosis. The next extract follows the first (though some intervening talk has been omitted). But notice how the doctor manages to avoid actually stating the child’s condition even as she makes her recommendation. 

What the doctor has done is to glide from a statement of the evidence (from the tests) to a recommendation for treatment, passing over actually naming the child’s condition. It falls to the mother (at line 9) to make explicit what has so far been implicit. Maynard has noted this pattern in his work on news delivery in mundane conversation (Maynard 2003). The news deliverer organises their hints at bad news in such a way that it is the recipient who is prompted actually to pronounce it. In ordinary social life that hinting has a set of implications which we might interpret as being to do with the complexities surrounding death and other taboo issues; in the clinic, it has all those, but also has more prosaic consequences as well. If the patient (or their representative, as in the case above) is the one who comes out with the news, it shows that he or she has been attending to what the doctor said, at least enough to work things out for themselves; it puts patient and doctor on something of an equal footing. Certainly it is more equal (or more equal-looking) than would be the case if the doctor simply pronounces the condition straight off.