When people talk about self-organization, the conversation often turns quickly to roles, responsibilities, decision-making autonomy, or agile ways of working. That makes sense. In companies, self-organization usually becomes visible when teams take on more responsibility or when decisions are no longer made exclusively through hierarchy.
But this perspective only tells part of the story. Self-organization is not primarily a method for modern collaboration. It describes a fundamental phenomenon of complex systems: order can emerge without a central authority fully planning or controlling it.
We can observe this in biological systems, ecosystems, social groups, and organizations. Wherever many elements interact, structures, patterns, and dynamics can develop that cannot be traced back to a single author or designer.
For organizations, this is a far-reaching insight. It challenges an assumption deeply embedded in traditional management thinking: that desired outcomes are mainly achieved by planning them, designing the right structures, and then controlling their implementation.
Complex systems do not behave quite so linearly. To understand why, it helps to look at five concepts: emergence, homeostasis, resilience and antifragility, evolution, and autopoiesis. Not as isolated technical terms, but as different perspectives on the same question:
How does organization emerge when it cannot be fully organized from the top down?
Emergence: Order No One Designed
A team can be described quite precisely on paper. There are roles, responsibilities, processes, goals, and perhaps even detailed rules for how decisions should be made. And yet, none of that tells us exactly how the team will actually work.
Over time, its own communication patterns begin to emerge. Certain people are repeatedly approached for specific questions even though they are not formally responsible. Others take on informal leadership roles. Some conflicts are addressed openly, while others are consistently avoided. Routines develop, shortcuts appear, and knowledge becomes distributed in particular ways.
No one has fully designed all of these patterns. They emerge from the ongoing interaction between the people involved.
This is what emergence describes.
Emergent properties appear at the level of the overall system even though they cannot be attributed to any single part of it. The behavior of a team therefore cannot be fully explained by looking at each team member individually. What matters just as much are their relationships, interactions, and the patterns that arise from them.
For organizations, this has an important consequence: not everything that makes up an organization can be shaped simply by designing structures. That does not make leadership or organizational design irrelevant. It changes the nature of design itself. Instead of assuming that a desired state can be produced directly, another question becomes more important:
What conditions make certain patterns more likely to emerge?
If, for example, an organization wants decisions to be made more locally, that decentralization cannot be created through role descriptions alone. It requires boundaries, information flows, and decision-making spaces in which decentralized action can actually develop and endure.
From this perspective, self-organization does not mean leaving organization to chance. It means recognizing that order is not created through planning alone.
Homeostasis: Why Organizations Can Be Remarkably Stable
If structures can emerge, another question immediately follows: Why do organizations not constantly change beyond recognition?
This is where homeostasis becomes useful.
The concept originates in biology and describes a system’s ability to keep certain conditions within a stable range even when its environment changes.
The human body, for example, constantly regulates its temperature. When it gets hot, we sweat. When it gets cold, the body responds differently. Stability is therefore not created by the absence of change. It is created through continuous adjustment. This perspective can also be applied to organizations.
Organizations develop routines, expectations, rules, decision-making paths, and informal conventions. These reduce the need to reassess every situation from scratch. Over time, people learn how decisions are usually made, which behaviors are expected, and what counts as an acceptable solution.
That creates stability — and stability is useful. Without it, collaboration would become extremely costly. Homeostasis becomes particularly interesting, however, when an organization wants to change. Many transformation initiatives encounter a familiar phenomenon: new roles are introduced, decision-making processes are redesigned, or levels of hierarchy are reduced. At first, a great deal seems to be moving. A few months later, however, everyday behavior looks surprisingly similar to what came before.
It is tempting to describe this simply as “resistance to change.” That explanation is often too narrow. Organizations contain many mechanisms that stabilize existing patterns. A team may be expected to make more decisions independently while significant budgets still require central approval. Responsibility may be distributed while career paths continue to reward traditional individual leadership. Mistakes may officially be reframed as learning opportunities while still being punished in practice.
In such cases, the old state does not necessarily return because people are secretly opposing change. Large parts of the system are simply still producing it. For self-organization, this is critical. It is not enough to add freedom. We also need to understand which existing structures repeatedly recreate the old state.
Resilience and Antifragility: Staying Capable When the Plan No Longer Works
As long as everything unfolds as expected, it is relatively easy to optimize an organization for efficiency. Things become more interesting when something happens that was never part of the original plan. A key person becomes unavailable. An important customer suddenly changes their requirements. New technology calls established processes into question. A market develops differently than expected.
This is where resilience becomes visible.
Resilient systems are able to deal with disruptions, adapt to changing conditions, and maintain or restore their essential functions. This reveals an interesting tension between efficiency and resilience.
Imagine an organization that has optimized its processes over many years. Everyone performs a highly specialized role, spare capacity has been reduced, and processes fit together with great precision. Under stable conditions, such a system can be extremely efficient.
But when one of its central assumptions changes, that same optimization may become a weakness. If knowledge is concentrated in only a few people, there is no redundancy, and decisions may only be made in a small number of places, unexpected events quickly create bottlenecks.
Things that may initially look inefficient from an optimization perspective — overlapping capabilities, decentralized decision-making, redundancy, or spare capacity — can be highly valuable from the perspective of resilience. Self-organization can create additional room for action here. If people are able to make decisions where relevant information actually arises, every unexpected situation does not first have to travel through the entire hierarchy.
Antifragility goes one step further.
The concept describes systems that do not merely withstand variation and appropriate levels of stress, but can under certain conditions gain from them. For organizations, an important distinction is necessary: not every form of pressure makes a system stronger. Permanently overloading people is not antifragility.
The concept becomes useful when an organization can turn disruption into learning. If an unexpected event leads to knowledge being distributed more broadly, decision-making paths being improved, or new options being developed, the system emerges from the experience with different capabilities than before.
Resilience therefore asks: How do we remain capable of acting despite disruption? Antifragility adds another question: How can disruption help us develop additional capabilities?
Evolution: Why Organizations Need Variation
If we accept that the future cannot be fully predicted, another fundamental problem appears: we cannot know with certainty which solution will be the right one for a future situation. Organizations often try to solve this problem through better planning. More data, deeper analysis, and more sophisticated decision-making processes are meant to help identify the best solution in advance.
That is useful — up to a point.In complex situations, knowledge often emerges only through action.
This is where evolution provides a useful model.
Evolutionary processes can be simplified into variation, selection, and retention. First, different variants emerge. Under particular conditions, some prove more successful than others. Successful variants are then retained or reproduced.
Applied to organizations, this means that new solutions do not arise only because someone designs the right answer in advance. They also arise because different approaches are allowed to be tested. That requires variation.I f every team must use exactly the same process, little variation can emerge. The same is true if every deviation needs central approval. If teams are allowed to try different approaches, by contrast, multiple solutions can develop.
Self-organization can therefore create space for experimentation. But variation alone is not enough.
Variation without selection does not automatically produce development. Organizations need to be able to observe what works, what does not, and what can be learned from the differences. They need feedback and criteria that help determine what should be retained, adapted, or discarded.
The key capability is therefore not knowing the perfect solution immediately. It is creating a system that can generate different solutions, observe their effects, and learn from them. Under uncertainty, this capability can matter more than the illusion of being able to predict every relevant development.
Autopoiesis: How Organizations Continuously Recreate Themselves
The concepts discussed so far lead to a deeper question: If an organization is more than an org chart, a collection of roles, and a set of processes, what is it actually made of?
This is where autopoiesis offers a particularly interesting perspective.
The concept was developed by Humberto Maturana and Francisco Varela to describe biological systems. It refers to systems that continuously produce and maintain themselves through their own processes. The idea was later applied to social systems in sociological systems theory. In the context of organizations, it opens up a different way of thinking about what an organization actually is.
An organization does not exist like a building that is constructed once and then simply remains in place. It is continuously reproduced through communication and decisions.
A decision creates conditions for further decisions. Previous experiences shape expectations about what kinds of behavior are sensible or acceptable. Rules are applied, interpreted, or changed, influencing later situations in turn. Communication repeatedly connects to earlier communication, and through this process the organization continuously reproduces the patterns by which it remains recognizable.
This helps explain why formal changes often have less impact than expected.
A company can remove layers of hierarchy and formally give teams more decision-making autonomy. But if people still wait for the former manager to decide when something important happens, if budget approval remains centralized, and if mistakes continue to carry negative consequences, much of the previous organization remains intact.
An organization is not produced by its org chart alone. What matters are the patterns that are continuously recreated in everyday action. This also changes the way we think about transformation. The question is no longer only:
What new structure do we want to introduce?
It also becomes:
Which decisions, expectations, and communication patterns continuously recreate our current organization — and what would have to change for different patterns to emerge? At this point, autopoiesis and homeostasis begin to connect. Organizations do not simply possess stable structures. Through their own activity, they continually help reproduce that stability.
Self-Organization Is Not a Method You Simply Introduce
Taken together, emergence, homeostasis, resilience, evolution, and autopoiesis create a very different picture of self-organization. Emergence shows that order can arise without being fully planned. Homeostasis helps explain why systems stabilize existing states and why change is often more difficult than a new structure on paper might suggest.
Resilience and antifragility draw attention to how systems deal with the unexpected and how they may learn from disruption. Evolution shows why adaptation requires variation — and why organizations need spaces in which different solutions are allowed to emerge in the first place.
Autopoiesis, finally, makes clear that an organization is not a finished object. It is continuously reproduced through decisions, communication, and expectations. Self-organization therefore becomes more than a question of autonomy or delegation.
It is not only about who is allowed to decide. It is about how order emerges, how it is maintained, and how it can change. Perhaps that is why the question “How do we introduce self-organization?” is not the most interesting one. A more useful question might be:
What conditions does an organization need for meaningful self-organization to emerge, stabilize, and continue to evolve?
Anyone who engages seriously with that question begins to see organizations differently — less as something that can be fully engineered, and more as complex systems whose development can be influenced without ever being completely controlled.