I wrote last time about controversy regarding null hypothesis significance testing (NHST) in psychology. NHST belongs to quantitative research. There is, however, a whole other type of research in the social sciences: qualitative. Qualitative research doesn't use numbers, which is why a lot of students prefer it. That is, until, they actually get to do it and realise it is much more difficult than they thought. Perhaps the most basic form of qualitative research is thematic analysis. This is where you, for example, interview someone and look for common themes in what they say. Strictly speaking it is more of a tool which other forms of qualitative analyses may use. There are many different forms and most of them are difficult to pronounce (Interpretative Phenomenological Analysis) but the one I want to talk about here is Conversation Analysis (CA).
To explain what CA is it is easier to begin by describing Discourse Analysis (DA). DA is an umbrella term for similar analyses but in essence, DA looks at how words create and reflect meaning and reality. So, for instance, describing immigrants as 'flooding' in to the country reflects a very different reality to simply saying they are 'arriving'. Words have power to shape discourse and this is especially true in the world of politics. CA, however, doesn't look at the meaning of words and what they create: it looks on words as action and at what they do. It breaks speech down into parts and examines where those parts come in a conversation, why, what they 'do' and what happens if they are absent.
Most of us are not aware of it but we all follow normative rules when we take part in a conversation (Stivers, Mondada & Steensig, 2011). We don't have to follow them and we often don't, but when we don't follow them the smooth progress of a conversation is disrupted and we have to work hard to get it back on track. What are these rules? There are too many rules in conversation to go into here in a blog post. However, conversation analysts have identified something called adjacency pairs (Goodwin & Heritage, 1990): a first pair part (FPP) followed by a second pair part (SPP). This is jargon but easy enough to recognise: it's any opening that requires a closing. So, when I say hello (FPP) you say hello (SPP). If I say sorry (FPP) or you say 'that's okay' (SPP). A phone ringing is a FPP which requires you to pick up (SPP); a sneeze is a FPP that requires a 'bless you' (SPP).
This is simple enough except adjacency pairs don't necessarily come so closely together. Nor is the SPP always provided. Often what happens, the FPP is given but a lot of work has to be done by both of you before the SPP is forthcoming. An example of this is where I tell you that I fell over. That would be the FPP. The SPP I want from you is likely to be sympathy. What will probably happen, though, is that if I tell you I fell over you'll want to know why, so instead of a SPP you'll begin a process of evaluation. Was I running? Did lots of people see? Did I hurt myself? Your SPP, if it ever comes, depends on these answers.
If a SPP is not given, the teller is caused something like distress. If I say sorry for being late and you don't respond I won't leave it there. I am likely to escalate my story by explaining my alarm didn't go off, or the train was delayed, or there had been an accident. If you still don't respond I will go on: 'I left early to allow myself time', 'I ran all the way' and so on. We do so much to elicit a SPP and if we never get it it can bother us for a very long time.
Given all this, I might want to pre-empt your evaluations (or lack of them) by something called prefacing. This is where I set the tone of the story and offer you a 'candidate stance' (Antaki, 2012). That is, the stance or opinion I want you to take. This is quick and easy to do and very effective.
'I fell over' doesn't say much at all.
'I went up to collect an award and fell over' does much more.
'I went up to collect an award and slipped on water' does even more still.
By the time I have completed my preface the candidate stance that I was embarrassed and injured and it was not my fault is well set up.
It's not always so simple, though. Candidate stances position the teller of a story (Wetherell, 1998). In the above example I am positioned as a blameless and reliable witness. Had I tried to absolve myself of responsibility for doing something dreadful by saying I was drunk, I am positioning myself as someone whose testimony might not be accurate. A witness to the event who had not been drinking is then positioned as the authority.
This battle for positioning between the teller of a story and the recipient is called 'epistemic primacy' (Stivers, 2011). That is, who has the right to knowledge? You might think it is obvious the teller does, but this is not always the case. If I tell you I saw a foul in a football match that should have been a penalty even if you didn't see the incident if you believe you know more about football than I do you will claim epistemic primacy for yourself. You might do so by asking me where I was sitting at the time, how much football I watch, which side I supported and so on. By the end of it, even though you were not there and I was, you could be in full charge of the story.
All this is just a very basic introduction to CA. It gets a lot more complicated! Conversation analysts scrutinise every nuance of speech, which turns out to be very predictable more often than not. And that is the point of it all. Think of all the interactions where it is very important it goes smoothly. If you call the police in an emergency you would hope it all goes quickly and efficiently. Well, CA can help with that. Doctor/patient interactions are another example. Think, though, of being a parent and being told by your child that something bad happened that day. How do you handle it? Do you respond with sympathy immediately or do you withhold sympathy and ask a series of evaluative questions first? CA suggests it's the latter. This gets even more important when you learn that children of evaluative parents tend to do better at school and score more highly on measures of self esteem.
The worst part of CA is that when you start to do it you actually become a worse listener. You start to recognise the 'rules' as they are followed or broken. You start to be able to predict what's coming next. If a conversation follows a 'textbook' route you'll smile to yourself. And all the while, you haven't really listened to a word that's been said!
Antaki, C. (2012). Affiliative and disaffiliative candidate understandings.Discourse Studies, 14(5), 531-547.
Goodwin, C., & Heritage, J. (1990). Conversation analysis. Annual review of anthropology, 283-307.
Stivers, T., Mondada, L., & Steensig, J. (Eds.). (2011). The morality of knowledge in conversation (Vol. 29). Cambridge University Press.
Wetherell, M. (1998). Positioning and interpretative repertoires: Conversation analysis and post-structuralism in dialogue. Discourse & Society, 9(3), 387-412.
