SenseMaker contours of narrative – how a culture might evolve, where a culture won’t shift

Published by Tony Quinlan on

Moving cultural beliefs and patterns is tough – the way that people behave generates micro-narratives and stories that then inform (constrain) how other people behave.  Feedback loops and patterns that can strengthen over time, making change seem impossible*.  In our work with SenseMaker, we see this in organisational cultures, in national values and beliefs and in excluded populations’ behaviours. And what we see is that the engineering approach to change – set a target, plan a route to get there, communicate clearly, introduce rewards and punishments – is fundamentally flawed. It’s hard, it’s resource-intensive and it is rarely sustainable without regular reinforcement.
Treating it as an ecological problem makes changes more sustainable and, often, lower cost. But that takes a different approach – understanding what values and behaviours are at the moment (disposition) and where they might or might not be inclined to go next (propensity)**. The theory behind it has been around for a little while – fitness landscapes based on micro-narratives collected in Cognitive Edge’s SenseMaker tools. In Narrate, we’re starting to dig into the practice and what it might actually mean.  Bear with us as we try to explain this…

Imagine a SenseMaker project is gathering multiple micro-narratives from babies and having these self-signified at the point of capture by the babies themselves***. One of the desired impacts is to improve the process of going to sleep.  Using SenseMaker Analyst, we can see the following contour map relating to this problem:

 So – our desired direction is to the right of the image – the kingdom of baby sleep, beloved by parents everywhere. As all scouts and mapreaders will recognise, the more closely packed the contour lines, the steeper the slope at that point on the map – in this case showing a deep valley on the left hand side of the contour map. In complexity terms, this could be seen as an attractor pattern of behaviour – the regular pattern into which it is easy to fall. And the steepness of the slope out indicates how hard (and resource-intensive) it would be to impose a change in that direction.

We know, however, that our own behaviours change depending on the role we’re playing or the circumstances around us. The same can be said for babies.  So now, let’s look at how this contour map changes in different situations:


These four graphs show the state of the baby’s disposition in four different situations:

  1. With the television on in the background
  2. In a state of exhaustion
  3. With a nappy that needs changing
  4. Where a favourite book (the Gruffalo) is clearly visible

In each of those situations, the contours look slightly – and usefully – different.  We can also see from the shape and closeness of the contours, the propensity of the baby towards changing.  Roughly, these arrows indicate likely possible paths from this point in time:

So we can see that, broadly:

  1. With a television on in the background, the likelihood is that baby is less inclined to go to sleep – the fitness landscape makes it easier to move to the left than to the right.
  2. In a state of exhaustion, the move towards sleep is easier than with the television on, but is unlikely to support a bedtime story – the fitness landscape makes it easier to move down and to the right.
  3. With a nappy that needs changing, there are tight contour lines on the left-hand side – very little about this landscape is going to shift anytime soon.
  4. With the Gruffalo on offer, there is the possibility for sleep to kick and no active resistance to a story. (Not much positive effect of one either, but at least it is not contra-indicated)

From all this, based on the SenseMaker evidence, we can say:

  • If a nappy needs changing, little shift in attitudes and behaviour is possible
  • If a television is on, any shift in attitude and behaviour is likely to be the oppositive of what we want
  • If the baby is exhausted, don’t bother trying with a story
  • The easiest path is, if the above three conditions are resolved, to make it plain that the Gruffalo is on offer

This, in turn, gives us an indication where to start addressing the problem/system.  Nappy, television, etc…

It’s only one of the powerful insights that come out of SenseMaker’s contour maps – but there is still the need for human sense-making both in interpreting the maps and in deciding which maps to explore.

But now consider the work that we actually do – looking at populations, employees, customers.  And imagine what the axes might be in those situations – engagement with democracy, customer focus, brand loyalty. Mapping these systems can give us deep insight into what’s impossible (or at least unlikely) and what is sustainably possible – and what we may need to address first.  It might also tell us that our current desired direction is impractical from where we are today – but help us to understand what is possible, knowing that our subsequent actions will shift values and behaviours and open up new opportunities as a result.

The effect therefore is to work with the current system, understanding where it’s possible to go and helping it to evolve in those directions. It’s not easy, but it’s a better, more sustainable and more effective route than the standard engineering-style approach. “Up the wooden hill to Bedfordshire” is hard work if the hill’s steep – easier to stroll across the rolling plain of doziness…

*And the feedback loops and narrative patterns mean that traditional change projects and communications may have an initial apparent effect, but this dissipates over time if the original patterns have not changed.

**This obviously has implications for the Nudge/Behavioural Economics ideas for which Thaler won the Nobel Prize yesterday. Rather than try different things to see where people can be pushed, first explore where they’re inclined to go (or not) and then design nudges accordingly – taking into account their current position and concerns.

***The images and SenseMaker® data here are from a real client project, but to maintain confidentiality, obviously the axes and the issues have been changed.  We won’t even go into the difficulties of collecting data from babies.


Steve Grey · 12 October 2017 at 11:16 am

The descriptions of these contour maps rarely touch on the Z-axis. What did you use for that?

    Tony Quinlan · 12 October 2017 at 1:07 pm

    It’s one of the things about these 2D contour maps that is better both from a theoretical perspective as well as from a practical one over the old 3d landscapes. The old 3D landscapes needed a Z-axis – usually stability (or a proxy thereof). These contour maps, in my understanding, create the contours based on the probability density of the dots – overlaying them on top of the XY plot. I remember Alicia Juarrero saying a long time ago that that was a more appropriate way of generating the landscape, so a) that’s good and b) it means we don’t need the stability proxy in the framework.

Comments are closed.