Know your self (organisation)

TLDR: Self organisation isn’t something you can teach, it’s an inherent property of the team and system. Coach teams to make use of it using constraints and attractors.

(if you want more of this kind of thing, sign up to my Substack where I'll be publishing in the future)

The Problem

 

The idea of self organisation has become pretty commonplace in the agile world. I’m going to tell you about some of the origins of the concept in cybernetics and complex dynamical systems theory, and argue that 

1. Self organisation means something more precise and different from how you may have thought of it.

2. If we don’t understand that definition we will make mistakes that get in the way of our agile teams rather than helping them

On an individual level, self organisation can be used as ‘get yourself organised’ in a kind of self helpy, ‘get your shit together’ kind of way. At the team level it often translates to ‘the team gets to choose how to work’. This usually comes with a value judgement as well, ie that it’s good for the team and the team’s agility that they get to choose how to work and therefore become self organising. This is at least how I understood it until recently. 

According to this definition, if I want to help a team with their self organisation, I need to help them get more organised, wrestle more power from management and generally be more autonomous. If I do that it’s good for everyone. 

I’m going to argue that this definition is missing a key understanding of how a team works and how to get the best out of them. 

Self Organisation is an inherent property of some systems  

In Physics there are certain contained scenarios in which many elements in a system interact with each other, and out of these interactions interesting properties emerge which are different from the sum of their parts. For example in a car, we have different components interacting, giving us a predictable thing we can use. The car-ness of it isn’t contained in any one component, it’s a property of the correctly functioning system as a whole. 

A complex dynamical system on the other hand is one where those parts interact in such dynamic ways and cross-cutting that it’s very difficult to predict exactly which specific properties will emerge or where. For example on a motorway the cars produce this thing called traffic, which is a property of each driver trying to get where they are going safely and quickly each dodging around each other, hopefully according to the rules of the road. It’s difficult to predict exactly how traffic will present itself day by day, but there are certain stable qualities that emerge. We’ll come back to this example later. 

Self organisation is defined in cybernetics as a property of a system which without any central control or design organises itself. ‘Organised’ here is defined in terms of order - a system is more ordered or organised if it takes less information to describe it. 

For example if there was just roads and no rules, and all the cars were going in different directions at different speeds however they like, it would take lots of information and descriptions to tell you about that system. We would have to say ‘one car is going south east in a straight line at 90mph, another north wiggly waggly going between 50 at 55mph, etc etc’. Lots of information is needed to describe disorganisation. 

On the other hand if all the cars are driving on the left (forgive the Anglo-centricity) in the same direction it takes less information to describe that, and it’s more organised or orderly. Similarly if everyone had to drive at 70mph exactly that would be more organised (at the expense of being more unsafe). I could describe a team as more organised if it makes all its key decisions in one meeting on a regular schedule rather than randomly as people think of them. This behaviour may or may not be helpful mind you, but it’s a kind of organisation.

What makes something self-organised is when that order comes about spontaneously from the basic rules of its components. Some elements of car driving are centrally organised and mandated by law, but the precise movements and behaviour of all the cars is self-organised in that it emerges from the decisions that each driver makes in their own context. 

Lots of types of system can be self organising; - ball bearings assembling into wires under a current, Conway’s Game of Life producing mind bendingly beautiful patterns, weather patterns. Self-organising systems generate a kind of stability and order by themselves. 

A human is a complex system made up of organs, organelles, cells, molecules and atoms. Each of these elements has their own highly ordered self-organising properties which express themselves as biological processes. At the cellular level, The simple instructions from DNA combined with the complex chemical interior and exterior of each cell results in self-organising behaviours which keep us alive and kicking. Especially in the case of our leg cells. Sorry. 

Notice that in the case of human, nobody teaches you to self-organise. You don’t have an organ coach who sits down with your organs and shows them how to influence each other to keep your body at the right temperature. Self-organisation is an inherent property of the human system at the body level of organisation. 

In the Embodied Mind, Varela Rosch and Thompson describe cognition as an emergent property from the self organising mind. Put simply (hopefully) - our minds are made up of lots of component elements; our brains, the rest of our bodies, the environment, our actions, and the interplay between all those things. It solves lots of the problems of cognitive science if we treat the mind as a self-organising biological and environmental system, out of which the phenomenon of mind emerges, just like the patterns in Conway’s game of life. This is the more modern alternative in Cognitive Science than, say, looking for a brain region that represents every object we have heard of, or a specific neuronal pathway for raising my hand. 

So too our teams and organisations are inherently self-organising. Human beings will arrange themselves to create order in systems at local and global levels whether instructured to or not. You can’t teach self organisation any more than you can teach a human to have eyes. They just do. Usually. 

The humans in your team are self-organising systems from which useful - and sometimes harmful - biological life-giving properties emerge. They display self organising behaviour as a result of their interactions with their environments, and the teams and company will be subject to the ways in which the individuals self organise. Some of these will be good and helpful towards the organisation’s aims (collaborative high functioning teams), and some will not (maybe toxic behaviours or a culture of evading responsibility). 

This challenges the idea that more self organisation for a team is always good. Lots of the bad stuff we want to get rid of comes from self organisation too. 

So what makes one system different from another? How can we use this knowledge with our teams? 

There are properties of the system which influence the types of self organisation that is possible in a system. Philosopher Alicia Juarerro (and others) refer to these as constraints. Constraints are the bounds within which a system operates which influence its behaviour. Constraints limit options within a system which has a significant impact on what is possible in that system. 

My parents got married in 1981 and just celebrated their 42nd anniversary. On that fateful day they agreed not to date anybody else and to limit lots of their attention to each other. They decided not to live just anywhere but to live together and to only have children with one another. They imposed these constraints on the system of their relationship. These in turn influenced various self organising behaviours from the two of them, and the people around them, eventually including from my sister and me. These self organising systems produced an emergent quality - the Susser family. 

Going back to our car analogy from earlier, some of the constraints operating on that system might be physical (the weather, the material of the road and the wheels), others might be imposed by authorities (the speed limit, the legal manoeuvres). These are what we call governing constraints, limiting the allowed moves. Another kind of constraint is at play called enabling constraints - those which force another kind of ordered dynamic into the system. 

On family trips to see our grandma in Nottingham when I was growing up we would drive by the Ratcliffe on Soar power plant to see the Ugly Chimneys, as we called them. We would slow down and wave at them because we didn’t have Sky TV at home and this brought us joy. No doubt the cars swerving around us was an emergent property of the enabling constraint of my family’s love of the ugly chimneys. 

Traffic policy makers don’t instruct all drivers wheel-turn by pedal-push how to become self organised in a traffic scenario. A driver has essentially infinite legal options about how to use the roads. What the government and authorities can do is tweak the constraints of the system and observe how that impacts the inevitable self organisation of the road system. Creating and enforcing a speed limit leads to a self-organisation which produces fewer high speed collisions and faster journeys, which I hope you agree is overall a good outcome.

If policy makers didn’t tweak those constraints, road users would still self-organise. They would just self-organise into other emergent structures which - assuming that the traffic policy makers do a good job - come at greater risk to life and limb. 

Whereas a weather system can’t be aware of itself and change its own constraints, humans can in fact do this. We are self-making, self-interested, self-observing organisms and as such can create our own destiny. We have the power to shape ourselves by shaping our environment. If I eat too much chocolate and it’s getting in the way of my goals to keep my teeth healthy and reduce the number of porsches my dentist can buy with my money, I can impose a constraint by implementing a rule that I never buy chocolate in my weekly shop and see if that allows me to self-organise to become an organism that eats less chocolate. Maybe it will lead to me stealing more chocolate from the office and scoffing it guiltily in the car park. Who knows? That’s something I have to find out. 

What does this mean for our teams?

This logic has an impact on our professional organisations at three levels. 

1. We can coach individuals to map, understand, challenge and - if needed - change the constraints operating on them to see what emerges. We can do the same with our teams. 

2. We can look at ways to help the team tweak the constraints that bind them, and sit alongside them as we witness how self-organisation within the constraints of their own making unfold. 

3. When working with leadership we can teach them this logic and expose the constraints that they are placing on teams, either consciously or unconsciously. 

One more element of a complex dynamical system worth mentioning is an attractor. These are elements in the system that bring other elements towards its behaviour. A delightful version is the video that went viral a few years ago of one guy dancing by himself at the bottom of a hill in view of a huge crowd at Sasquatch festival in 2009. He dances exuberantly by himself for about a minute, then one guy joins, then eventually another, and before long there’s a huge joyful spontaneous dance party with people running down the hill to join in. In this case that man was an attractor, pulling behaviour to become more like his own. 

Certain practices can be attractors in our organisations. For example, I’ve worked in so many organisations where leadership are happy for the teams to self-direct and choose how to work, but the teams are under the impression that management wants them to use story pointing. This can lead to a weird kind of self-organisation where the attractor of story pointing subjugates many other behaviours under its centre of gravity, leading to people taking part in practices that everyone thinks are stupid and non value-adding. That would be an example of an attractor that is producing a negative self organising behaviour. 

The point is that if we are looking for ways to improve we should constrain ourselves to deal in constraints rather than imagining that we can teach the team to self-organise directly. Though we can’t teach a team to self-organise, we can teach a team how to manage their own constraints and attractors. Though we can’t control the outcome of changing constraints, we can help a team self-reflect and become better at handling those outcomes, noticing them and tweaking the constraints and attractors. 

So at the end of this long piece what are the take-aways? 

• Teach teams *about* self organisation, even if you can’t teach them to self organise 

• Help teams to make judgements about the constraints and attractors that are impacting their self organisation for better or worse

• Teach senior leadership about this when they are seeking change, and especially if they have a big change initiative on the horizon. For anything like that to be successful it has to be built on understanding of how self organisation works.