A complex system never really settles down for good. Even without any big event pushing it, the people, organisations, and conditions inside it keep shifting on their own, because they are always exchanging influence with everything around them. A system you understood clearly last year may already be behaving differently today, without anything dramatic having happened at all.
This is related to adaptation and domains of stability, though the emphasis is different again. Adaptation is about a system reacting specifically to something you did. Domains of stability is about a system settling into one resting place or another. Change over time is a simpler and, in some ways, more humbling point. Even without any intervention from you at all, and even after a system has genuinely settled into a stable state, the wider context around it keeps moving, which means that the stable state itself can eventually shift underneath you. If you want to look at either of those related ideas on their own, we wrote about them separately in Adaptation: What It Actually Means in Systems and Domains of Stability: What They Actually Mean.
We described a version of this before, without naming it change over time. In How Do We Know Our Work Is Actually Working: Theory of Change, we wrote about a decade long review at the Tribal Health Initiative. Health outcomes had genuinely improved since the programme began. But an internal review a decade in found that the underlying poverty driving migration and family stress had not moved at all, a reality the original plan had never accounted for. Not because anyone had gotten it wrong at the start, but because the surrounding situation itself had kept moving in the years since. A theory of change built on how things worked a decade earlier was, by then, describing a system that no longer quite existed.
Where You'll See It
The same pattern shows up in systems that have nothing to do with each other:
A partnership shifting when one partner changes. A group of organisations working together can find their whole collaboration change direction simply because one partner updates its own internal policy, even if nothing else about the shared goal has changed at all.
An ecosystem moving through stages on its own. Bare land left alone does not stay bare. It fills with fast growing weeds first, then shrubs, then eventually trees, each stage changing the conditions enough to make the next stage possible, with nobody managing the sequence at all.
River pollution building up around cities and villages. A river can absorb a certain amount of waste and still recover on its own. Households, factories, and towns along its banks each add a little more over the years, and nothing looks alarming for a long time.
Then one year the same river cannot clean itself the way it always used to, not because of any single spill, but because of everything that was added slowly, over decades.
The One Thing People Get Wrong About It
It is easy to assume that once a system has been carefully mapped or understood, that understanding stays accurate for as long as the programme runs. Change over time says something less comfortable. The system you mapped honestly two years ago may already be behaving differently today, not because your original mapping was wrong, but because the system itself has kept moving since then.
A map is a photograph of a moving system, taken on one particular day. The question worth returning to periodically is not whether your understanding was right. It is whether it still is.
Key Takeaways
Do not assume a system has reached a stable state just because it looks calm right now. What looks settled today may simply be between changes, not finished changing.
Do not rely on a system staying the way you last understood it. The people, organisations, and conditions inside it keep shifting on their own, whether or not you are intervening at all.
Revisiting your understanding of a system is not a sign that the original work was wrong. It is simply what working with something that never fully stops changing requires.
The next time a plan or a map still feels accurate after a long stretch of time, it might be worth checking whether that is because nothing has changed, or because nobody has looked closely in a while.
What are the complex systems you have been working with? Tell us in the comments!
Related reading: Adaptation: What It Actually Means in Systems · Domains of Stability: What They Actually Mean · How Do We Know Our Work Is Actually Working: Theory of Change