Systems Thinking
Foundations / Systems ThinkingSee the whole, not the parts. Behaviour lives in interconnections; feedback loops, emergence, and mental models explain why local fixes go wrong.
systemSystemElements interconnected to produce a behaviour. Swapping a part rarely changes behaviour — it lives in the connections.
loopsFeedback loopsReinforcing loops amplify (growth/collapse); balancing loops stabilize. Delays cause overshoot and surprise.
emergenceEmergenceWhole-system behaviour unpredictable from any part. A traffic jam is not a property of one car.
modelsMental modelsSimplified beliefs about how the system works. Make them explicit; revise against feedback. Disagreement = clashing models.
⚑Before fixing a symptom, find the loop. Ask what structure produces the behaviour, and intervene there rather than on the symptom.
Symptom vs structureHeavy approvals cut incidents, then push devs to batch risky big releases — incidents return. Fix the structure, not the symptom.
Meta-skillCharacteristics (scalability) are emergent; hold each design as a current mental model, revised against feedback.
systemsfeedback-loopsemergencemental-modelsthe-whole