Cynefin and Corona

Cynefin (pro­nounced kuh-NEV-in) is a “Welsh word mean­ing habi­tat,haunt, acquaint­ed, famil­iar. Snow­den (who invent­ed the Cynefin-Frame­work – edi­tor’s note) uses the term to refer to the idea that we all have con­nec­tions, such as trib­al, reli­gious and geo­graph­i­cal, of which we may not be aware.” (Wikipedia)

The Five Domains of the Cynefin-Frameworks

Essen­tial­ly, the Cynefin frame­work sug­gests that sit­u­a­tions or con­texts can be of very dif­fer­ent nature and there­fore require very dif­fer­ent approach­es to deal with them. David Snow­den dis­tin­guish­es there­fore five domains: sim­ple or obvi­ous, com­pli­cat­ed, com­plex, chaot­ic and disorder.

For sim­ple or obvi­ous sit­u­a­tions there are check­lists, process­es and proven solu­tions. Here it is actu­al­ly only a mat­ter of sens­ing the sit­u­a­tion, cat­e­go­riz­ing it and thus reach­ing into the right draw­er and react­ing with the appro­pri­ate solu­tion. If, for exam­ple, a warn­ing light comes on in a car, the man­u­al usu­al­ly con­tains instruc­tions on what to do and how to do it (e.g. refill oil).

Things get com­pli­cat­ed when the man­u­al no longer offers advice. For instance, if you have checked and refilled the oil in the engine and the warn­ing light is still on. Because there is no longer a sim­ple solu­tion, you go to an expert in the garage. With enough knowl­edge and expe­ri­ence, the expert can ana­lyze the prob­lem and choose the most promis­ing one from var­i­ous options for action.

To man­age a sys­tem effec­tive­ly, you might focus on the inter­ac­tions of the parts rather than their behav­ior tak­en separately.

Rus­sel Ackoff

Some­thing com­pli­cat­ed can be dis­as­sem­bled by experts and can be under­stood through its com­po­nents. The whole is the sum of its parts. This is no longer true for com­plex sit­u­a­tions. Our brain is a com­pli­cat­ed net­work of neu­rons and the bio­chem­i­cal process­es in it can be under­stood by experts. But my thoughts in this net­work can­not be pre­dict­ed by ana­lyz­ing the neu­rons. Com­plex sys­tems are always more than the sum of the parts, they are the prod­uct of their inter­ac­tions. There­fore, rela­tion­ships of cause and effect in com­plex sys­tems and sit­u­a­tions can­not be explored by decom­po­si­tion and analy­sis, but must be empir­i­cal­ly researched in order to make them vis­i­ble (and thus at least par­tial­ly shift the prob­lem into the com­pli­cat­ed domain).

Final­ly, a chaot­ic sit­u­a­tion is char­ac­terised by the fact that cause and effect is not appar­ent and yet action must often be tak­en under extreme time pres­sure. The ter­ror­ist attack on the World Trade Cen­ter on Sep­tem­ber 11, 2001, was one exam­ple. In such a sit­u­a­tion, the first step is always action in oder to gain some sta­bil­i­ty and thus to trans­form it step by step into a com­plex situation.

In all chaos there is a cos­mos, in all dis­or­der a secret order.

Carl Jung

And if it is not even clear which of the four domains men­tioned is pre­dom­i­nant at the moment, David Snow­den calls it dis­or­der. The first thing to do in a sit­u­a­tion like this is to deter­mine the right domain (sim­ple, com­pli­cat­ed, com­plex or chaot­ic) at least for spe­cif­ic parts of it and to act accordingly.

Classification of the Corona Pandemic

So in which domain of the Cynefin frame­work are we cur­rent­ly in the midst of the SARS-CoV­‑2 coro­n­avirus pan­dem­ic? The sit­u­a­tion is cer­tain­ly any­thing but sim­ple or obvi­ous, and it is not com­pli­cat­ed either, because in many cas­es cause and effect are still unclear even to experts. So we are mov­ing some­where between com­plex and chaot­ic, or rather we are on the way from chaot­ic back to complex.

The first reac­tions in most coun­tries were the very abrupt restric­tion of pub­lic life up to a com­plete lock­down. These mea­sures ini­tial­ly helped to reduce chaos and cre­ate more sta­bil­i­ty. Now that we have arrived through these reac­tions in the com­plex domain, the next step must be to bet­ter under­stand the causal rela­tion­ships through empir­i­cal research. What hin­ders or pro­motes the spread of the virus in soci­ety? Which mea­sures are effec­tive and which are rather ineffective?

In order to achieve this as effec­tive­ly and quick­ly as pos­si­ble, diver­si­ty and dis­sent are need­ed. Dif­fer­ent approach­es and mea­sures in the indi­vid­ual coun­tries, fed­er­al states or even cities are not a mis­take, but a good oppor­tu­ni­ty to learn faster togeth­er. But this requires two things. On the one hand, we have to allow and acknowl­edge this diver­si­ty in the vari­a­tion of mea­sures with­in cer­tain con­straints, and on the oth­er hand, we need a stan­dard yard­stick, as far as pos­si­ble, on how the effects of these mea­sures are to be eval­u­at­ed and compared.

In my per­cep­tion, we cur­rent­ly have deficits in both respects. On the one hand, giv­en the diver­si­ty, experts and lay­men alike get embroiled in heat­ed debates about the one and only right mea­sure and why the oth­er mea­sures tend to endan­ger lives and are there­fore irre­spon­si­ble. On the oth­er hand, how­ev­er, the key fig­ures are also con­stant­ly chang­ing. In the begin­ning, much atten­tion was paid to the time it took for case num­bers to dou­ble, then to deaths, then to the num­ber of free beds in inten­sive care units (which admit­ted­ly had to be painstak­ing­ly record­ed first), and now it is the repro­duc­tion num­ber R, which absolute­ly must be kept well below 1.

It would there­fore be par­tic­u­lar­ly help­ful in this phase to define the essen­tial objec­tives in a com­pre­hen­si­ble man­ner and to ensure that they are con­tin­u­ous­ly record­ed with the small­est pos­si­ble delay and in the best pos­si­ble qual­i­ty and that they are acces­si­ble to all. Unfor­tu­nate­ly, we in Ger­many still have a lot of catch­ing up to do in terms of dig­i­tal­iza­tion (fax!), which is painful for us now, because the feed­back cycles are con­sid­er­ably longer as a result.