Self-Organisation, Self-Management, Self-Determination and Autonomy
There is a confusion of language to describe teams which don’t need external direction in order to deliver effectively: self-organising, self-managing, self-determining, autonomous… I noticed this when writing up my thesis, but it’s not really relevant there and thought this might be interesting to other folks. Another trigger to write this was that a friend of mine recently wrote, “Scrum is built on self-management”. Except, of course, it wasn’t build on self-management until 2020, when the The Scrum Guide changed from:
Scrum Teams are self-organizing and cross-functional. Self-organizing teams choose how best to accomplish their work, rather than being directed by others outside the team. — (Scrum Guide 2017, p6)
to:
Scrum Teams are cross-functional, meaning the members have all the skills necessary to create value each Sprint. They are also self-managing, meaning they internally decide who does what, when, and how. — (Scrum Guide 2020, p5)
At the time of this change, Dave West of Scrum.org noted:
The reality is that both terms come with baggage and history. For some self-organization only describes how the team is organized and how they organize the work. For others, it describes empowering the team to do whatever is required to deliver a valuable increment. — Dave West
I acknowledge that I am late to this party; Willem-Jan Ageling also wrote about this back in 2020, and highlighted that many people think the terms are interchangeable. He also highlights that “self-managing” is a shift away from the language used in Principles Behind the Agile Manifesto, and in A Scrum Book, which both refer to self-organising.
Here’s how I think they differ, and why, and the language I prefer to use. I’d be super interested to hear your views on this because language is not fixed and is always changing. Do you use other words I’ve not considered?
Self-Organising
The original word in the domain, which comes from complexity theory. These days, this word is mainly used by physics to describe spontaneous re-arrangement of an open system (ie. a system which can exchange information, energy and/or matter with the environment) which is far from equilibrium, and which has a maximum of one macro-constraint on it. For example, convection cells (the bubbles in simmering and boiling water) form when water heated from below creates a temperature, and therefore pressure, differential, creating a system which is not in equilibrium. The formation of cells is constrained by the amount of energy coming in, and the distribution of that energy in the system; if the temperature equilibrium is restored, the water will stop producing convection cells.
These spontaneous changes are unpredictably predictable (a convection cell will form when the temperature differential reaches a critical point, but you cannot predict whether it will rotate clockwise or anticlockwise in advance). These spontaneous changes are often described as phase shifts in the system, for example, water moving from liquid state to gaseous state (steam), and can exist in both states at the same time (double-point). The final phase change of heating water is when it is rapidly boiling and chaotic. Climate change is an example of what happens when a phase change occurs in an open system, as global temperatures rise, the weather systems self-organise into patterns to cope with the increasing amounts of energy in the system, so we see more rain, more storms, more wind, more extreme weather events.
In my experience, this is a description of a team that is able to reconfigure themselves to switch between different types of work demand (for example, a team which can switch to “production incident mode” and back when it is resolved), analogous to phase changes in heated water. Such teams are able to switch between a limited number of modes which they have successfully handled in the past, but when presented with a new challenge might end up in chaos like a rolling boil.
Self-Managing
This word is used more in biology, where it describes (open) systems which maintain dynamic homeostasis (ie. maintenance of a given status quo) to ensure conditions compatible with life, the classic example here being temperature self-regulation, or, at a lower level, protein production in cells. Such biological systems are generally subject to multiple constraints, and those constraints may arise from, and contribute to, the components of the system (for example, the number of actions available to a cell is limited by the content of the genome, the raw materials available in the form of nutrients, oxygen and the like). This is known as closure of constraints where the constraints mutually depend on each other for their continued maintenance and viability. Biological systems can have reciprocal levels of mutuality and closure as well, meaning a system at one level (eg. organism) can contribute to the collapse of a system at a lower level (eg. cell) or at a higher level (species) by not being able to support the replacement and repair of the constraints necessary for the processes, and this goes in both directions. Systems (eg. cells) are therefore dependent on and depended on by other systems (eg. the organism) for mutual maintenance.
The changes in these systems are generally predictably predictable, which is to say that if you get hot, you will start sweating (unless you are a royal who fought in the Falklands war, obviously), and that the sweating will get rid of the heat as a result of evaporation. Since this is an open system, the environment is important here: air pressure, humidity and temperature all affect the ability of this system to be effective. High humidity means that the sweat doesn’t evaporate; a temperature that is too high means that you can’t sweat enough to reduce your body temperature sufficiently. An inability of self-management leads to the collapse of the systems ability to maintain coherence, and leads to death, whether at the cellular or organism level.
Self-managing teams have the components and constraints in place which allow regulation and maintenance. For example, external pressure on a team to deliver features can result in a them not dealing with technical debt, which can lead to system non-viability over time; if the team is unable to regulate that external pressure, then it is not capable of self-management. In addition, a team which switches from one mode of operation to another (eg. from new product delivery to product maintenance and support) may become non-viable because they don’t have the capabilities necessary.
Self-Determination
Self-determination introduces a concept of agency (the ability to act) into self-managing (open) systems, which is to say that the system has a purpose or intent, or to use a ‘long’ Greek word which I am told endears me to others, telos. The telos defines the norms of the system, and the norms of the system reinforce the telos, creating a mutually reinforcing dependency between them. The norms constrain the behaviours in the system, but as the telos can change, so can the norms. This is considered a form of agency as it adapts to achieve its own baseline, and different systems of the same type in the same metasystem (eg. cells in an organism), and systems of the same type in different metasystems (eg. cells in different organisms) can show diversity in their constraints whilst achieving the same goals. (To be clear, this doesn’t require sentience or choice by the system; cells don’t have individual consciousness, although of course humans acting within a purposive system do. In addition, I am not making a teleological argument here whereby a system must have purpose X because it has particular norms; these can be teased apart).
In addition, as changes occur, the telos can change. For example, as a haematopoietic stem cell becomes a B-cell it changes its telos from being a very generalised, immature cell to being a cell that prototypes and presents antibodies against pathogens ; when a B-cell comes up with an antibody that is effective against that pathogen, it becomes a memory cell or a plasma cell and its telos changes to either hang around and detect future infections by that same pathogen, or it becomes producing as many of those antibodies as possible as quickly as possible to mount an attack against a current infection. As a result, the systems in these cells turn on and off in order to adapt to the new telos and maintain new norms (ie. patterns of constraints).
This seems much more appropriate to me for teams; if the teams purpose changes, then the norms in the team need to be able to change with it. For example, this is why agilists often speak of separating the WHAT and the HOW of work: if the work to be done (the what) changes, then it’s likely that different tactics, patterns and behavioural norms (the how) will be needed, and they are they people best positioned to make that assessment. If the team — whether that’s a software engineering team, a product management team, or a senior leadership team — is unable to adapt how they work to a changing environment, then they will be doing the wrong thing, however well they’re doing it.
Thus, an effective self-determining team requires them to be able to adapt their norms to a changing purpose. An organisation which has only one way of working is by definition not particularly agile; an organisation which overly constrains norms likely inhibits teams from being able to achieve their purpose.
Autonomy
Autonomy arises when the system is able to select and refine their own telos. This is of course constrained by the systems around it: if you’re a cell in a liver, then deciding your purpose is to create urine (a kidney function) is at best misguided, and at worst maladaptive and dangerous.
B-cells, mentioned earlier, select their own telos, depending on what antigen fragments are presented to them by other immune cells, thus the selection is mediated by environmental constraints (if a B-cell never gets exposed to a particular antigen it cannot become a producer of antibodies against the pathogen from which it came).
Effective agile teams need to be able to determine and change their telos because they are able to detect changes in the context in which they are working, whether that’s noticing more query time-outs so prioritising database performance, or detecting consumer priority changes a product line and adapting an organisational strategy to better meet what is being demanded. As a result, they are then able to adapt their norms. However, that purpose may no longer be relevant, so appropriate norms (such a reflecting on purpose and norms) can allow the purpose to change to maintain the viability of the system.
This means that autonomous teams not only need to be able to adapt their norms in a self-determining manner, they also need to be able to change their purpose. They need have norms which enable feedback loops to identify when that needs to happen, they need to be trusted to set the right purpose for their context, and they need to have the power to change their norms to best meet the telos. A team which cannot adapt their purpose to changing conditions will rapidly become non-viable, as the norms and behaviours will no longer being delivering contextual value.
Conclusion
The language we use shapes the way we think about what is happening, and considering the teams we work with can allow us to develop realistic assessments of their capabilities. By thinking about how we think about teams, we can help ensure that the development work we are doing with them — helping to develop capabilities and skills — are appropriate to the teams, their contexts and the demands for adaptability.
