Control and Complexity
Carsten Busch
Safety Mythologist and Historian. The "Indiana Jones of Safety". Grumpy Old Safety Professional.
Reading Jake Chapman’s nice little book “System Failure” on systems thinking (you can download a free version of the first edition), I noticed a fantastic metaphor that I would like to share. One of the threads in Chapman’s recommendable work is that we need to approach and understand complex systems (like organisations) differently than simple, linear, mechanic systems.
Chapman illustrates this with an example, which he takes from Paul Plsek, who in turn draws on Richard Dawkins. Find it on page 51 (and other places) of “System Failure”, and also in the October 2001 issue of the NHS magazine Leading Edge (*). This example compares linear cause and effect to complexity by comparing throwing a stone to throwing a bird.
When you throw a stone, you can predict with a rather good degree of certainty where it is going to land. The stone is governed by the Laws of Mechanics and those are rather linear. If you know the stone’s weight, the force and angle of your throw, friction, gravity and the like, you can calculate where it will land, within a reasonable margin.
When you throw a living bird, it is in principle governed by the same Laws of Physics. Yet even when you know the bird’s weight, air resistance, the force of your throw, side wind and what not, it is very, very unlikely that you manage to predict where the bird is going to land.
Unlike the stone (**), the bird is a living creature. The bird is a complex adaptive system and it will adjust to the challenges it encounters. We can calculate and predict as much as we like, but the bird will follow the path through the air that it wants and lands where it wants. It might need a few moments to handle the initial shock, so for the first tiny part of the path through the air it may actually look like our prediction is correct. After that, however, the bird’s reflexes kick in, take over and it will do something completely different from that parabolic trajectory that we calculated.
We can of course try to make the bird following our calculated trajectory. We can attempt to force linearity onto this complex adaptive system, like when we wrap the bird in duct-tape, possibly weighing it a bit down with a rock. Doing so will increase the accuracy of our calculated, predicted trajectory significantly, but it comes at a cost. It completely destroys the capability of our bird.
In this example, it should be pretty obvious that the bird should NOT become more stone-like in its behaviour (***). Yet, this is exactly what many organisations, managers, and safety practitioners try to do when they force their reductionist, linear models onto situations that they are not suitable for. Stones have to be handled by one approach, birds by another. If we want the bird to land on a certain spot, we should not reduce its mobility and use force, but rather give it a good reason to land where we want, for example by placing some food, or a bird handler. Likewise, complex systems should not be forced through mechanistic, linear approaches. They may seem to give short-term effects; remember that first bit where our bird actually followed the trajectory? In the long run, however, it will be more valuable to exploit the capacities of the system and let it achieve the goals. The bird can fly by itself, so why use force at all.
(*) By pure chance, I also came across this video with the same case.
(**) Actually, in Michael Crichton’s novel “The Andromeda Strain” you can find an argument that we cannot be entire sure about this. Maybe the stone is alive, and has an entirely different perception of time and speed? For our purposes, we will regard them as innate objects, however.
(***) This may remind us of the old wisdom that if theory and reality do not agree, it is rarely reality that is wrong…
---
Carsten Busch is a self-declared Safety Mythologist and author of the well-received book Safety Myth 101. This book collects 123 (and then some) of Safety Myths. Crisp and compact discussions address weaknesses of conventional safety ‘wisdom’ and give suggestions for alternative approaches and improvement.
https://www.mindtherisk.com/the-book
Health & safety | Human Factors | Consultancy & research | Train, facilitate, coach | SME and usability | Content simplification |
7 年I wonder if the "Wiser at Work" presentation has missed the point? Maybe we shouldn't be encouraging the bird to go a particular way by throwing bird seed - maybe the bird knows better than we know where it should go?
Health & safety | Human Factors | Consultancy & research | Train, facilitate, coach | SME and usability | Content simplification |
7 年It's a great analogy. The bizarre thing about Richard Dawkins is how such sense and such nonsense can come out of the same head. Perhaps he is also a complex adaptive system...
Heart-Centred Leadership / L&D / Communications Specialist ERS | Risk Management, Cert NLP Master Pract
7 年Wow - great story! Thank you for sharing Carsten Busch
President at Erickson Associates
7 年Carsten, an applicable metaphor. There is far too much linear thinking in safety. Even though necessary, it should be supplemented with the less seemingly tangible, yet equally relevant, so-called "softer" factors instrumental in the level of safety performance. Of course a system is more than a sum of its parts since the parts themselves interact. If one has ever done a fault tree analysis this is clearly evident. Therefore, taking any one part of a system gives us no answers, the same as taking one point of data is meaningless. To some of us this is very clear. To others, not so much.
Safety Mythologist and Historian. The "Indiana Jones of Safety". Grumpy Old Safety Professional.
7 年Those who think that complexity can be tackled by a reductionist approach: it may seem that this gives results (especially short term, just think of the trajectory of our bird), but do consider that a system, and especially a complex system is more than the sum of its parts. So when you take it apart and zoom in on the parts, you will only get limited information about the system. Say you have a team of three persons (A, B and C) that is not functioning well. You can look at the three persons individually (the reductionist approach), but that ignores most of the information, like the interactions A-B, A-C and B-C. And of course the trio of A-B-C. And that interactions can be different in different situations and for different subjects. If you want to know about the team, you have to observe it as a whole, not as a couple of separate parts.