THE SCIENCE OF CONVERSATION

Conversation is our theme – see our essay, Conversation as the New Model for Qualitative Research, or the shorter Powerpoint version, with pictures, here.

But, are we qualitative researchers ‘conversation experts’, or is what we do a pale imitation of what ‘real’ conversation experts do? There is a growing field called Conversation Analysis (CA), which was the latest subject of the AQR’s excellent In Depth. (This post initiated by a review commissioned by AQR last week).

Why should any of us bother to learn about CA? As Ailean Mills says, we should ask ourselves if our methods actually help or hinder ‘naturally occurring conversation’. For example, the number of questions we typically ask, the discussion guide and the role of the moderator, these may hinder, not help: ‘the moderator… can determine and distort the discourse of a discussion from the very outset …The choice of language, and its structure, will be quickly mimicked by participants’.

The focus in CA is on ‘naturally occurring conversation’ and all the examples quoted come from real-life examples of conversations, such as a doctor’s interaction with a mother in a clinic and a telephone conversation between two friends. Most groups and depths, of course, are far from naturally occurring!

Which raises two key questions:
1.    Should we all be incorporating more ethnographic methods, getting out there and listening to what real people really say (without their realizing we are listening in)? If we did this, though, how long would it take before we learned anything about Pedigree Chum (or any other brand)?
2.    Are only naturally occurring conversations ‘real conversations’? Can an ‘artificially created’ conversation (ie all focus groups and interviews) not contain real conversational elements and therefore valuable insights?

What is a ‘real conversation’?
It occurs to me (and I am biased) that parts of an artificially-produced conversation can be just as revealing as a naturally-occurring one, in the same way as a farm-produced strawberry can be just as nutritious as a wild one. The obvious benefit of the farm production method, of course, is a massively greater yield (far more strawberries/insights for your money).

To extend the horticultural metaphor further, we have to be very conscious of/careful about the ‘production’ methods we use. You can spray plants with all kinds of harmful stuff and cram them into boxes and into freezers, in the interests of a better commercial return. But the quality of the end product suffers. So when we are producing ‘artificially created’ conversations, how to ensure maximum quality?

CA, as written about by Professor Celia Kitzinger and Merran Toerien, uses a particular notation, as for example in this data fragment, taken from a telephone conversation between two friends, Belle and Fanny, about the death of someone they both know.

[DA:2:10, from Schegloff (1988: 443)]
01  Bel:  ... I, I-I had something (.) terrible t'tell you.=
02        =So [uh:  ]
03  Fan:      [How t]errible [is it.]
04  Bel:                     [.hhhhh]
05        (.)
06  Bel:  Uh: ez worse it could be:.
07        (0.7)
08  Fan:  W'y'mean Ida?
09        (.)
10  Bel:  Uh yah.hh=
11  Fan:  Wud she do die:?=
12  Bel:  =Mm:hm,

Could we not extract more insight by a closer attention to not only what people are saying but also the way they say it? And, are there better ways to revisit discussions than by reading so-called verbatim transcripts?

Of course, despite using these analysis methods, an artificially-created conversation is not the same thing as a naturally-occurring conversation. I would argue that for the majority of commercial qualitative research projects for private sector clients, artificially-created conversations are more useful, more insightful and ten times more productive than naturally-occurring ones.

Ethnographic methods
But can we not also learn from the practice of following more naturally-arising conversations?  We are supposed to be experts in what ‘people-as-consumers’ say and do, not just in what ‘people-as-respondents’ tell us.

This would mean a greater pursuit of qualitative ethnomethodologies, which I think is a bit of a trend at the moment. More and more researchers are (apparently) offering clients consumer diaries, mobile phone texting of encounters with brands, on-street/in situ conversations etc.

However, much of what passes for ‘ethnographic’ can be challenged, as Caroline Hayter points out in her excellent article - there is no real agreement as to what constitutes an ethnographic enquiry.

This surely is something we, as qualitative research practitioners, working with the AQR, can help to set up. Ethnography is a whole different ball game, as we are discovering through our work with EverydayLives. A client’s agenda gives you a structure and a set of priorities that it is hard to wean yourself off.

Doing ethnographic research, properly so-called, opens up new avenues for qualitative researchers and starts to take us into new domains. It would be great if the AQR followed up on the principles and practices of CA in the future, maybe on some of the more practical aspects of how to do it. CA represents a good stretch for our discipline and (IMHO) makes life more interesting.

Maybe it is up to us to take clients with us.

But that is a whole other conversation.

Comments

[Paul Feldwick]

I can't get my head round the idea some conversations are 'natural' and others not. I would go with Patricia Shaw's idea that all conversations are continual negotiations and contructions of relationships and sensemaking. So any participant can dominate or direct what's happening at any one time, but there's also the potential for this to change. In most commercial research I would guess there is a fairly strong drive by the 'moderator' to hold on to this sense of control as much as possible, and that will restrict the freedom of the conversation to evolve. Rather than stick with these notions of naturalness, it might be more interesting for researchers to reflect on issues of power and control in what they do - and the fact that they are part of the reality they are trying to study.

[Kevin]

thanks for this thoughtful comment, Paul ... I was trying to distinguish between CA's 'naturally arising' conversations ie happen in real life without a moderator and 40 quid involved, and the typical qually's more 'pointed' (more mercenary?) conversations, thinking that there IS, despite everything, a considerable overlap, cos we humans converse like we breathe, it defines us and sustains us, and although our clients set the parameters, whereas CA practitioners have no such constraints, still the less we 'moderate', the more we participate, the better.

[Jon Hindmarsh]

For those interested in more on CA there are some additional references that might be of particular value and that relate to issues you raise in your post Kevin:

Claudia Puchta and Jonathan Potter have used CA to analyse the interactional organisation of focus groups in market research and include implications for moderators:
'Focus Group Practice' (2004, Sage publications)

Also Colin Clark and colleagues have undertaken a series of CA studies to unpack sales encounters (from street markets to tele-shopping to telesales to B2B sales and most recently studies of browsing in retail stores). I'd be happy to forward on the full references to anyone who is interested but their first major contribution was their book 'The Hard Sell' (1995, Harper Collins).

[Greg Rowland]

Jon, thank you for this contribution, much appreciated, will read references ...

what do you think CA can make available to commercial qual research?

[Jon Hindmarsh]

(Too long - sorry!)
As stated in the ‘In Depth’ piece, some companies are starting to use CA in amongst developments with ethnography, mainly technology companies like Xerox, Intel, Microsoft, etc., to inform the design of new systems (speech recognition systems, video conferencing systems, audio guides, etc.). More generally my feeling is that there are at least 4 areas where CA could be particularly useful in commercial environments:
1. Understanding the skills of the focus group moderator to help in the training of moderators and the analysis of focus group data, which is the issue picked up by Aileen Mills (see the Puchta and Potter book for some initial pointers on that).
2. Understanding verbal persuasion. Max Atkinson wrote a wonderful book on political speechmaking (‘Our Masters’ Voices’, 1984) which drew heavily on CA and gives a real sense of what works as persuasive language and how it can be seen to be persuasive. Using his analysis he trained a novice to get standing ovation at a Liberal party conference (for a ‘World in Action’ TV programme) and then started a company for training communication skills: www.speaking.co.uk. Others have used similar approaches to study persuasive talk in meetings, management presentations, sales talk and the like. These sorts of studies could be applied from anything from preparing business pitches to developing radio and TV advertising.
3. Understanding the power of the sequential organisation of conversation to grasp a range of practical problems: how to build rapport without appearing sycophantic (ask for the customers’ assessment of something and then agree by demonstrating your agreement rather than simply claiming it), how to get someone’s name without asking for it (give your own in a greeting and wait for the response!), how to keep a conversation alive (ask another question), how to anticipate disagreement (as discussed in the In depth piece)… etc, etc. All of these can have useful practical value in specific scenarios.
4. However at its heart CA has an interest in the organisation of social interaction in everyday life. I have referred here to how existing studies could be used to inform practice. However I think that there is value in it as a method to be adopted in business to understand all sorts of issues of relevance to clients: how conversations, debates and discussions emerge is relation to TV advertising; how participants *structure* disagreements and debate in focus groups; how conversations between salespeople and customers emerge at point of sale (and the implications of different opening turns); how consumers structure objections to sales work (and the implications of different responses to those objections); and so on and so forth.
There are some initial thoughts. It's difficult without specific commercial examples, nevertheless it’s something that I would be keen to explore further.