The way of thinking that explains why the fix did not work. It treats a situation as stocks, flows and feedback, and asks what arrangement of those would produce exactly the result you are complaining about.
The behaviour comes from the structure, so look at the structure
Systems thinking describes a situation as parts connected by feedback. A result that keeps recurring is read as the output of a structure, not of a person or an event. Its central tools are the loop and the delay: the consequence that comes back to change its own cause, and the effect that arrives too late to inform the decision that caused it. It is strongest on problems that have resisted repeated fixes, and weakest when the structure is not what is actually stuck.
Systems thinking
In ordinary life
The traffic that comes back
A town widens its main road to cure congestion. For a year it works. Then the congestion returns, as bad as before, with more cars on a wider road. Nobody made a mistake in the engineering. The wider road made driving easier, so more people drove, until driving was exactly as slow as it had been.
Systems thinking is the habit of expecting that. Forrester called the behaviour of social systems counterintuitive, because the obvious intervention often feeds the loop that caused the problem. Senge put it more bluntly: today’s problems come from yesterday’s solutions.
Outside work it shows up in small ways. A household that argues about who does the dishes sees a pattern instead of a person. The faster someone cleans up, the less anyone else notices the mess, and the more the cleaning falls to them. Nobody decided that. The arrangement did.
Meadows’ list of leverage points is the practical gift. Numbers and targets are the weakest place to intervene, and the one everyone reaches for first. Rules, information flows and goals are stronger. The mindset the system rests on is stronger still, and hardest to shift.
The educational experience
The loop behind the homework
Every page follows the same school on the same kind of day. The school has decided to increase students’ use of generative AI, and the plan is due for review. The five people and the situation stay fixed. Only the way of thinking changes.
All five, interconnected
One day 9 beats
The five threads as they were written, interleaved, so the hand-offs between them are visible. Pick a person to follow their day on its own. The people are illustrative, and the same five appear on every way’s page.
07:15
Rob, home from a night shift
Rob finds Nadia’s laptop open on a finished essay she wrote in twenty minutes. He is uneasy and cannot say why. The essay is better than anything he could have written at fifteen.
08:30
Mr Osei, before the review meeting
The adoption dashboard shows eighty per cent of students using the approved tool weekly. The target is ninety. His draft recommendation is more training for the departments below the line.
09:40
Ms Lim, Year 10 history
She sets an in-class question on the causes of the First World War, the same topic as last week’s homework. The homework essays were strong. The in-class answers are thin. The gap is wider than last year.
10:20
Nadia and Ms Lim, after the lesson
Ms Lim asks Nadia how she wrote the homework. Nadia tells her plainly. Homework is marked on the finished essay, the tool produces good finished essays, and she has four other subjects due that night.
12:50
The planning meeting
Ms Lim draws a loop on the whiteboard. More AI use raises homework marks. Higher marks make the homework look like it is working. So more homework of the same kind is set, which makes AI use more worthwhile. Understanding sits outside the loop entirely, measured only in class, weeks later.
13:15
Mr Osei, same meeting
He sees that his adoption target sits on the loop too. Every department pushed towards ninety per cent adds to it. The dashboard is measuring the thing the loop produces most easily.
15:30
Priya, in the regional office
Reading Mr Osei’s revised note, she recognises the loop. She has three other plans on her desk with adoption targets. The ministry’s own guidance asked schools to report adoption, because adoption was easy to count.
18:40
Nadia and Rob, at dinner
Rob asks about the essay. Nadia explains the arithmetic of her evening: five subjects, three hours. Rob had been ready to argue about cheating. He ends up asking which subject could do with less homework.
21:00
Ms Lim, marking
She rewrites next week’s homework as preparation for an in-class essay, marked on the in-class version. It is a small change to one rule. It moves the mark to where the understanding is.
Nobody in this day behaved badly. Nadia responded sensibly to how she was marked. Mr Osei responded sensibly to his target. The loop was built out of reasonable responses, which is why exhortation would not have broken it. Changing one rule about what gets marked did.
Implications
What this way of thinking does to teaching and learning
The same five dimensions for all four ways. This is the part that can be tabulated. The two narratives above are the part that can be argued with.
What it notices
Patterns over time, and the structures that produce them. A systems thinker hears "we tried that and it made things worse" as data about a feedback loop, not as bad luck. It notices delays, especially the ones between an intervention and its cost, and it notices when a measure has become a target and started changing the behaviour it was meant to observe.
What it cannot see
Individual people, most of the time. A loop diagram abstracts away the student who is the exception, and can make a real injustice look like a neutral structural effect. It is also weak on the moment of action. A good map of the system tells you where the leverage is and says very little about what to do on Monday.
And the boundary is a choice it tends to forget it made. Checkland and Churchman both insist that where you draw the edge of a system decides what the system is. A map of the school that stops at the gate will never find the loop that runs through the family’s evening.
What it asks of teachers and students
Patience with diagrams that look wrong for a while. Students need to learn that an arrow is a claim about causation that can be argued with, and that a loop needs a direction and a sign before it means anything. Teachers need to hold back the explanation long enough for students to find the loop themselves.
Good work draws the loop, names the delay and predicts what a proposed change would do before anyone tries it. Weak work draws everything connected to everything and calls it a system.
Where the values sit
In the boundary and the goal. Deciding what is inside the system and what the system is for are value judgements dressed as modelling decisions. Ulrich’s critical systems heuristics exists to make those choices explicit, by asking who the system serves, who decides, and who is affected but not involved.
How it gets hollowed out
The loop poster. A diagram drawn once at a retreat, admired for its complexity and never used to predict anything. The hollow version treats "it’s complex" as a conclusion instead of the start of an analysis. Meadows would say the map is not the point. The point is to find the few places where a small change shifts the pattern.
Reading
Where it comes from
Published works only. Start with the first on the list.
Meadows, D. H. (2008). Thinking in Systems: A Primer. Chelsea Green.
Stocks, flows, feedback and the leverage points. The best single introduction.
Forrester, J. W. (1971). Counterintuitive behavior of social systems. Technology Review, 73(3), 52–68.
Why the obvious intervention in a social system so often makes things worse.
Senge, P. M. (1990). The Fifth Discipline: The Art and Practice of the Learning Organization. Doubleday.
Systems thinking brought into organisations, with the archetypes that recur in schools.
Checkland, P. (1981). Systems Thinking, Systems Practice. Wiley.
Soft systems methodology. Systems as ways of seeing a situation, and the boundary as a choice.
Churchman, C. W. (1968). The Systems Approach. Delacorte.
The argument that every system boundary carries a judgement about who and what counts.