Dave Snowden

Counter narratives

RSS Feed


If you move around government and intelligence circles, or touch them tangentially (as I do) from time to time you can’t escape this concept. The idea is simple: there is some narrative you don’t like, it might be drug runners creating a popular culture that makes crime aspiration, it might be radicalisation in an inner city area or it may be romanticization of alcohol consumption amongst teenagers. There are a lot of candidates. Having identified that which is not desired, the cry goes up to create a counter-narrative that will destroy the undesired. It’s almost like they are looking for an anti-matter charge that will match and destroy matter that they don’t like.

The problem is that narrative is a little more complex than that and critically you are not dealing with a single narrative, but a pattern in the flow of multiple narratives. Any dominant narrative which is visible to the authorities using conventional scanning methods is well established and its underlying themes are highly resilient. Add to that the simple fact that if I want to believe something then its easy to only pay attention to those aspects of the counter-narrative that support the way we currently feel. Tacking a negative story directly (as any parent of teenagers know) is a mistake, you generally make the undesirable, desirable.

I sometimes think that its best to think of dominant narratives (the things people want to counter) as a flock of birds or a school of fish rather than as a concrete entity. Each of the stories gains purpose by interaction of other stories, they follow a form of Boyds algorithm: follow the next story, match its speed, avoid contradiction. Flocks are resilient, they don’t have leaders and you can’t counter them directly.

Now this links to issues around causality for which I owe a few back posts. Food poisoning post my visit to Rome last weekend took me out for four days and the backlog of email and papers/book chapters has hit so I am even more behind than usual. Now there is a key difference we need to make between teleology (explanation of phenomena by the purpose they serve rather than by postulated causes) and teleonomy (the quality of apparent purposefulness and goal direction that derive from the evolutionary history of the system). Now its a distinction that has been around for a long time (remember Aristotle’s four causes) and Kant’s determination of teleology as necessary principles weakened a wider view. With a growing understanding of complex adaptive systems its become a more vital (in several senses of that world) distinction. And by the way I need to acknowledge Max Boisot for the reminder of this distinction. I staggered from my sick bed to spend an hour or so with him before moving onto Parsifal on St David’s Day.

Now the problem is that teleology dominates western thinking and analysis. So when experts start to look at narratives, especially when they use semantic tools with all their limitations, they tend to clump things at too high a level of granularity. They talk about the dominant narrative as if it was intentional and/or designed. They attempt to deal with that causal perception as if it can be countered by rational argument or offering of directly linked inducements. In effect the dominant narrative of myth structure is seen as a thing that can be analysed, understood and countered.


In my work and that of my colleagues we have taken a different approach; for apparent purposefulness in that definition read retrospective coherence in complexity theory. Along with premature convergence (coming too quickly to a solution this is one of the two great perils of strategy in a complex world. In order to understand and to change things we need to change the flows of finely grained narratives something that can only be done experimentally after some mapping of the territory.

The chart to the right shows a real case of mapping. This is from a UK City, the vertical dimension indicates a focus on revenge, the y-axis is the degree to which the community is closed to new influences and the x-axis measures the strength in which upbringing is held to influence behaviour. Now to be clear this is based on the mass capture of micro-narrative using SenseMaker®, with the material signified or interpreted at the point of origin by those who told the stories. Now I could talk about this chart for ages, but at its heart it shows high levels of volatility where upbringing is less importance, in those contexts the possibility is or a rapid and un-anticipatable switch in the dominant narrative. Change here for bad or good is easy, unless upbringing is important at which point the landscape stabilises.

I hope this makes the point. The self-intrpretation of large volumes of micro-narrative provides objectivity in contrast with the inevitable cultural or political bias of expert interpretation. Where things are volatile small changes can produce major change. Where the landscape is stable investment in change is probably a mistake. This type of mapping allows us to understand and (if we capture continuously) monitor the flow of narrative within a community and construct and test safe-to-fail interventions. When you are dealing with telenomy you are dealing with evolutionary potential rather than cause and effect, with influence rather than direction. That means that change is a matter of starting to stimulate the formation of alternative not counter narratives; one properly such stimulation is done in parallel often with contradictory initiatives. Having spend time recently with people involved in transmedia this is also where they work. A film has a beginning, a middle and and end, to be satisfying it is teleological. The surrounding games, spin offs etc create a world of many possibilities that can be explored by participants.

It doesn’t matter if its counter-terrorism of cultural change. We need to starting mapping, experimenting and evolving rather than designing and analysing. We know there is a beautiful grain of sand out there but its on a beach, its a weak signal and it will only be discovered by interaction. Its also worth remembering that each discovery is unique in space and time. How often does a small intervention produce massive change (think of events in the middle east) but they don’t repeat. We see that when people try and copy good ideas cross-context they generally fail. What matters is tight monitoring and rapid response. The trouble is governments treat things as complicated, they sense-analyse-respond when they should probe-sense-respond. However you don’t probe at random, you map and test your probes for coherence against the map.

Thanks as ever to Gaping Void for the cartoon, but I wish they would provide high res images again.