What accumulates?
Inventory, cash, trust, skill, population, backlog and reputation are states carried from one moment to the next.
Side 07
Systems
Science
A study of behavior that emerges from relationships rather than isolated parts. Stocks accumulate, flows change them, feedback closes loops, delays distort timing, and interventions often create consequences somewhere other than where attention began.
structure→interaction→behavior over time→intervention
05system primitives
02feedback types
06archetypes
07Side
See the structure first.
A system map is not a list of components. It records what accumulates, what changes those accumulations, what information returns, where delays sit and what boundary defines the model.
What changes the stock?
Hiring and attrition, sales and returns, learning and forgetting, inflow and outflow alter accumulated states.
What comes back around?
An effect eventually influences one of its own causes, amplifying or counteracting the original movement.
What arrives late?
Production, hiring, perception, learning and policy responses often act after the conditions that triggered them have changed.
What did the model exclude?
Every map is partial. A boundary is a deliberate analytical choice, not a claim that the outside world does not matter.
Stock-flow identitynext stock = current stock + inflows − outflows
Events are snapshots. Systems thinking asks for the movie.
The useful unit of observation is often behavior over time: what rose, fell, oscillated, saturated, overshot or recovered?
Loops generate behavior.
Feedback is causal circularity. The distinction between reinforcing and balancing loops explains much of the behavior that initially appears mysterious.
Reinforcing loop · R
A change produces effects that push the original variable further in the same direction.
adoptionnetwork valueword of mouthnew adoptionmore value
Signature
Growth or decline that feeds on itself until another constraint becomes dominant.
Balancing loop · B
A change produces effects that oppose the original movement and pull the system toward a target or constraint.
temperature gapheatingroom warmsgap shrinksheating falls
Signature
Goal seeking, stabilization, resistance to change or oscillation when delays are substantial.
reinforcementcreates momentumconstraintscreate limitsdelayscreate surprise
Recognize behavior over time.
Different structures leave different temporal fingerprints. The pattern is a clue to the loop architecture underneath.
Compounding
A reinforcing loop dominates: growth adds to the base that produces further growth.
Growth meets a limit
Reinforcing growth starts strongly, then a balancing constraint progressively takes control.
Gap closes
A balancing process adjusts action in proportion to the distance from a target.
Correction arrives late
A balancing loop with delay repeatedly overshoots and undershoots its target.
Momentum outruns capacity
Growth continues beyond a sustainable level because limits are perceived or acted on too late.
The supporting stock erodes
Demand on a resource exceeds regeneration long enough to damage the system’s future capacity.
Recurring structures, recurring mistakes.
Archetypes are reusable hypotheses. They are useful as diagnostic prompts, not labels to force onto every situation.
01
Limits to growth
A reinforcing engine succeeds until a balancing constraint becomes binding. Look for the limiting factor before pushing the growth loop harder.
02
Fixes that fail
A quick correction relieves the symptom while delayed side effects recreate or worsen the original problem.
03
Shifting the burden
A symptomatic solution becomes easier to repeat while the capacity for a fundamental solution weakens.
04
Tragedy of the commons
Individually rational use of a shared resource accumulates into collective depletion.
05
Success to the successful
Early advantage attracts resources, producing stronger results and further resource concentration.
06
Escalation
Each actor responds to the other’s move, creating a mutually reinforcing contest around a relative target.
Where can the system actually move?
Interventions differ in depth. Changing a parameter is easier than changing the information, rules or goals that generate repeated behavior.
Parameters
Numbers, thresholds, prices, staffing levels, quotas. Visible and often easy to change; sometimes powerful, often temporary.
Buffers & capacity
Inventory, slack, reserves and time horizons. These affect how shocks travel through the system.
Flow structure
Physical or procedural pathways that determine how material, work or people move.
Delays
The time between signal, decision, action and consequence. Shortening or recognizing delays can prevent overcorrection.
Information
Who sees what, when and with what fidelity. New information flows can alter behavior without changing formal authority.
Rules
Incentives, permissions, constraints and decision rights that shape local choices.
Goals
What the system is optimizing for. A new goal can reorganize lower-level behavior.
Mental models
The assumptions used to define the problem, the boundary and what counts as success.
The ladder is adapted conceptually from Donella Meadows’ work on leverage points; it is compressed here into study categories rather than reproduced as her original twelve-point list.
System mapping drills.
Start with behavior, then propose structure. The objective is not to draw a complicated diagram; it is to discover a better hypothesis.
A team keeps adding staff, yet delivery time barely improves.
Map backlog as a stock. Add hiring as an inflow to capacity, but also onboarding load as a temporary drain on experienced staff. Add coordination cost as team size grows. Ask which balancing loops are cancelling the intended capacity gain.
A marketplace grows quickly, then quality deteriorates.
Map users, supply and trust as separate stocks. Growth may reinforce liquidity while verification capacity lags. Poor matches increase complaints, reduce trust and eventually slow both sides of the market.
A recurring emergency is solved faster every month but happens more often.
Test “shifting the burden.” The symptomatic response may be improving while investment in root-cause prevention falls because the emergency capability makes recurrence tolerable.
A policy change creates no visible effect for six months, then appears to overshoot.
Look for implementation, behavioral and measurement delays. A balancing intervention can seem ineffective early, invite additional intervention, then deliver both waves of effect after the system has already moved.
Thinking in SystemsDonella H. Meadows · systems literacy
Business DynamicsJohn D. Sterman · system dynamics modeling
Industrial DynamicsJay W. Forrester · foundational system dynamics
General System TheoryLudwig von Bertalanffy · systems foundations