Author: Manu Steens
In this blog I write my personal idea, not that of any organization.
A first reading for inspiration from the Cynefin framework in the booklet “The Cynefin-Book – An introduction to complexity and the Cynefin framework” gives inspiration to think about the crisis of covid19.
Contents
Cynefin Framework
Let me first explain the Cynefin framework itself, the theoretical framework that tries to create an order in types of systems, such as the world.
A first division is made between ordered systems and un-ordered systems .
An ordered system is strongly bounded, the behavior is very predictable. The causality is clear from experience, or is based on a virtuous analysis. The first case is a simple system , the second a complicated (but understandable) system .
We cannot establish causality in an un-ordered system. Some of these systems are stable, where the boundaries evolve over time, and so does the behavior. This evolution occurs through the interaction of its components. Evolution cannot therefore be called situational (i.e. caused purely by external factors) but dispositional: caused by internal, often unknown and changeable, dependencies and interactions. Thus, they are predestined to evolve in a predestined direction. But causality can only be proven retrospectively (in hindsight). These stable systems are the complex systems .
But some systems are not stable, they have little or no limitations or limitations, the behavior is or seems random. These are chaotic systems.
Finally, there are the systems that we have yet to classify, and those are the disturbed systems: we don’t know what they are yet.
How to solve them
The simple systems, we know them: “been there, done that”. They are known knowns. An unambiguous solution is possible for an unambiguous problem. The solution is given by SCR: sense, categorise, respond. We have best practices.
We no longer know what complicated systems are. They are known unknowns. But we can analyze it in great detail and then we can predict things about it. The approach is then SAR: sense, analysis and respond. We have good practices.
The complex systems are those of unkown unknowns. We need to conduct some experiments and hope they will guide us in the right direction. There are many hypotheses, without being able to call them wrong or right. So the approach is PSR: probe, sense and respond. Here we have newly emerging practices.
And the chaotic systems are the unknowable unknowns. The system requires immediate action, but we have no idea what the appropriate actions are. So we tackle it with ASR: act (do something), sense and respond. Here we have new but perhaps transient practices because they may only be usable once.
Finally, the desturbed systems, which we have yet to explore to classify them.
When a situation occurs
If a situation comes to us, the advice is: first differentiate which type of system it is. If we don’t know anything about it, or it is ‘forgotten’ matter and it seems something completely new, it is better to explore it first, bearing in mind that it can be a worst case system, so according to ordered or un-ordered systems a complicated or chaotic situation. When we do this, we avoid the risk of oversimplification in advance.
Covid 19
How does this fit with Covid19?
It was known from the first investigations that it was a corona-like virus. There was some experience with that. As a result, it appeared to be an ordered system. It soon became clear, however, that this particular corona was not as well known as originally thought. It was not a “simple cold”. The world with covid19 thus quickly became an un-ordered system from an ordered system. So we had to act quickly. In Philippe De Backer’s book “En nu is het oorlog” (“And now it’s war”) it appears that one took the the ASR approachvery quickly. And that was a good thing. The first wave was systematically dealt with severely in a radical way. There was a lockdown, and they contained the wave.
The province of Antwerp tackled an almost second wave with strict measures. So fast, in fact, that many dared to ask why it was necessary. For a while it seemed that the world under covid19 would become a complex system or even a complicated and ordered system. Been there done that? Existing techniques yielded new weapons: several types of vaccines saw the light of day. There was talk of “the light at the end of the tunnel”. Freedom was in sight. A chaotic system would be pushed back by science to a complicated system, where vaccinations are at the heart of it.
A new threat: groupthink
And a new danger threatens the battlefield: “groupthink”. By estimating it, it can be prevented.
Groupthink is an issue typical of complicated systems. Notwithstanding a great truth that lies behind the wisdom of the masses, as an average of a group of experts in the field. But then with standpoints that first make them independent of each other. This is to prevent them from influencing each other (too much). It is difficult to do this day-to-day, so an effective approach is needed against it, as well as to improve the scanning of information. To tackle this issue, it is best to involve different groups of experts, each from different areas of expertise, so that they can keep a ‘look of wonder’ about the other areas of expertise.
Questions
There are no stupid questions in such a group. But one must ask them. A question could concern the composition of the GEMS itself. After all, in addition to infectiologists, virologists and biostatisticians, are anthropologists, sociologists, behavioral experts, psychologists and weather experts also needed in an advisory body? The question occurred to me for two reasons.
- The disease always finds a way out in a new variant with which, when we return from a complicated system that seemed ordered thanks to the vaccines, to a chaotic system, which seems un-ordered by the arrival of a new deadly variant, but above all:
- The disease attacks through the people who are not ill at the time of the attack, but through their behaviors and habits threaten to expose themselves to the disease. The difference in people’s behavior during the different seasons determines the contacts they make. These behaviors thus form the channels through which the disease can spread.
Knowing the behavior
Knowing this behavior of the different target groups with a sufficiently small resolution with regard to the evolution in previous disease waves could provide better parameters for statistical models concerning those target groups and between them. Even though one can never see them as a model with a purely predictive value, it could provide an insight into the behavior of the past, so that one can create more refined expectations for the nearest future.
If these experts would already cooperate in the proposals and advice of an advisory body, this is not clear: I never see them in the news reports. I never hear any mention of their contributions.
Since the system on which a chaotic system must be dealt with is one by the acronym ASR, we can never assume that we know what is going to happen, even with the most sophisticated statistical models available at the moment. Therefore, as long as someone can make a coherent argument, their idea is valuable.
But with that comes the question: where does such a model lead? What will the approach to the current crisis lead to? Are they trying to beat the disease? Stop the spread? Or does it make more sense to work with a model in which one tries to control the disease?
Value of Statistical Indicators
Essential in a controlling mechanism is that one does not only look at the direct statistical measurements such as the Rt value, the number of sick people in the hospitals, the number of sick people in the ICU, the number of cases in the schools and the like. And that per province and over time. These are ‘direct’ indicators , the figures that we want low, except for the vaccination rates, which they want to see rise. By directly influencing behaviour.
If one accepts that corona will not just be defeated, but that we will have to serve a time in this prison called Earth, until the disease has adapted itself to its host, one could start in the fringe with ‘indirect’ indicators. By this I mean indicators with which one can indirectly influence people’s behavior, for which we have to rely on the expertise of anthropologists, sociologists, psychologists, together with weather experts and the experiences of the authorities with, among other things, tourism. One such indicator that one could track could be a “net good behavior promoter score”. One might take such a philosophy from a firm like Apple, which measures a “net promoter score” (NPS) to determine satisfaction from the customer’s perspective.
What this means for covid 19
For Covid19, this means that with such an indirect measurement one could know the involvement within the different target groups, which can indirectly measure the effectiveness of the press conferences. After some time, one can plot these numbers in an evolution, and see how it evolves, but also how it compares to the infections in the different target groups. By also targeting these indirect indicators, in addition to an attempt to stop the disease, an attempt could also be made to rather control the disease than directly to stop it. And that until the virus has previously been weakened for its host, or until a vaccination technique emerges that allows for a single vaccination that offers lifelong protection if that might be possible. The additional advantage of such a controlling effect could be that people are prepared for the idea of a one-time vaccination, which can then be more readily accepted.
Conclusion
Even if it has not been proven that such an approach would work in a short period of time, I think that in addition to trying to stop the crisis, we should also create every opportunity to try to control the crisis. But it is a lot of trying with ‘can’ and ‘could’ and ‘maybe’ one after the other. This approach would mean a radically different and additional approach to the traditional approach, in which one tries to reduce an un-ordered system to an ordered system on the basis of purely rational arguments. With which one tries to work on the behavior of the ‘homo rationalis’. With an indirect approach one can also work on those target groups where one has more to do with the ‘homo irrationalis’.