The Art of Conversation has been on a sabbatical for a few months. The Art of Silence. It’s been refreshing. And it’s also good to be back.
Loads has been happening in research while we've been off air, especially in relation to science and technology. There is huge interest in all things behavioural, which to a quallie with ethno leanings (like me) is all good. Well, it’s mostly good.
The thing is, I’ve also noticed a tendency for commentators to describe what I do (qualitative research) in terms I don’t recognize, then to denounce it for having these characteristics.
Detractors claim or imply that talking with people is worse than useless. It’s unreliable and non-predictive. People cannot be trusted. And it’s so-o last century. There’s an app for that now.
The assault on talking with people comes from advocates of brain science (BS), big data (BD) and behavioural economics (BE). It’s like the attack of the killer Bs. And there is plenty of it about. Especially the BS.
There is no doubt the research industry has been over-reliant on reported behaviour, which is why I welcome new approaches. But I fear that there are some risks heading our way from the perfect storm of technology and ‘New Behaviourism’.
When I started in research, we used to do research about people. Now, it's becoming unfashionable even to do research with people. Instead, given the rise of New Behaviourism, are we starting to do research against people?
Take technology. At the last MR Summit the key words from the expert panel were: datafication, microsegmentation, geolocation tracking, sentiment analysis, neural coding and gamification. And only one of these is made up.
Technology, we are told, offers a better way of understanding behaviour by being able to predict it. Sensors already track buyer demographics at the shelf. Now Tesco will start scanning our faces as we enter their stores. Or should that be if we enter.
Some have even suggested that advances in technology and computer speed mean that market research is dead. Well, nobody's said that before.
The problem with the increasing mechanization of research is that it is reductive. If you focus only on behaviour without attempting to understand context and intention, you are stripping out an essential part of the mix, the messy, unclear parts.
As Sheila Keegan has said, all this mechanization may just be an attempt to make people more 'controllable'; turning them into data and making them more tidy and convenient to analyse and monetise.
But human behaviour cannot be 'literalised' or adequately translated into bite-sized chunks of data. Research needs to explore and understand the whole experience of human beings, complete with contradictions, exaggerations and omissions.
This means continuing to have good conversations with the right people and using observation and other methods, rather than only using remote measurement, big data or other proxy research tools.
But New Behaviourists aren’t having that. A leading thinker recently wrote in ESOMAR's Research World that ‘traditional research' (note the framing) has been getting it wrong for ages, by talking to people. Such a practice is to modern behavioural science as blood-letting is to medical science. Ouch.
Mark Earls (for it is he) is an exemplary writer, thinker and speaker and all-round good chap. He is advancing his argument (Herd Theory) by first softening the opposition. Putting in a few early bouncers, as it were.
Mark does this using the principle of the Unreliable Narrator, (or sometimes unreliable witness, or poor witness). It is a central conceit and a delivery worth examining, since it is itself, in my view, unreliable. A no-ball.
The argument goes that people cannot be relied upon to give an accurate or reliable account of their own behaviour and for that reason there is no point in seeking their account.
“We are all unreliable witnesses to our own lives”
There are three things wrong with this, as applied to research. First of all, it assumes that we take what people say as gospel, as having some direct link to their behaviour. But who on earth does that? We all know that everybody is selective and reactive, exaggerates and fabricates. Knowing that enables you to start to deal with ‘narrator bias’.
Secondly, the unreliable narrator comes from literature, a created world, as a partial or selective account, a story. The value of stories is not whether they relate to facts on the ground, the-world-as-all-that-is-the-case, but how they relate to our shared, inter-subjective experience of the world. In other words, any narrator is ‘unreliable’ (ie partial or subjective). That is inherent in the idea of narration.
Thirdly, the New Behaviourist attack sometimes takes the form of people being a poor or unreliable witnesses and cite evidence from police interviews etc to say that people’s accounts of their experience are susceptible to influence. But we are talking about buying toothpaste and not witnessing a crime. It’s a brand extension, nobody died. And the job of the researcher is to encourage the person to be a better witness of their own experience. This can be done, but can take time and is not easy. It makes for useful training courses but does not make for interesting conference speeches.
But it relies on an authentic relationship between researcher and research participant, something which Roy Langmaid discusses comprehensively in his blog, the Langmaid Practice.
So the fact that we are, as I would put it, partial narrators of our own lives is prime territory for the kind of research I do, rather than being a reason to ignore it or not to do it.
I have been watching a wall near our offices over recent months, which is a site for one of Candy Chang’s ‘Before I Die’ installations. Every time I looked, people were reading it or adding to it.
Of course, many of the comments are not serious. But what struck me about this social art intervention is how it tapped into some fundamental human needs and desires. Such as the need to make sense of things, the desire to make a difference, the desire to leave one’s mark, the desire to explore one’s values, the desire to poke fun at social art installations. Not forgetting our basic, human curiosity.
What machine will measure these things? What can big data tell us about our dreams?
What does this tell us? Maybe that being 'unreliable' narrators is part of what makes us human. That we are not only all all storytellers and all want to make something out of our own experiences and our shared experiences (and these are linked).
It tells us that engaging with people, with the whole person, not a stripped out, stratified segment of a person, is essential in research. Talking with people, done well and for the right reasons, is still one of the best ways in research to understand why people do what they do.
More on stories to follow. Shorter stories, next time.