Thinking in Systems Summary

One Paragraph Summary The world is made up of systems. A company, a soccer team, your body, everything is a system. When we learn how systems function, we will give us a better understanding of the world itself. In Thinking in Systems the author, Donella Meadows, shows us what a system is, how systems function, and how we can use them to improve our problem-solving skills.

Thinking in Systems Summary

Thinking in Systems—A Primer
A book by Donella H. Meadows

  By writing the summary, I learned the following three things:

  • We cannot fully understand systems
  • A systems' purpose is deduced from its behavior
  • To understand a system, we need to make a model

Key Ideas

You can drive a system crazy by muddying its information streams.―Donella H. Meadows

Everything we see is a result of a system

Systems are all around us. Well, not only around us, we are made out of various systems ourselves. Most of these systems run within us without us noticing it.

In this summary, we will learn what systems are and how they work. You will also learn why feedback is essential for systems and how they can break.

What is a system?

We use the word system quite frequently, but have you ever asked yourself what it means? Well, don't worry, we will do this now.

The world itself is a system, and its constituents are subsystems.

Every system is made up of elements that are connected and serve a specific purpose. For example, a soccer team is a system with a shared mission of scoring goals. In this system, the players are the elements that are connected by relationships.

Another example of a system is a forest. A tree in this forest is an element in the forest system. But the tree itself is yet another system, a subsystem of the woods.

Elements of a system do not need to be tangible. They can be amorphous too. For example, a leaf of a tree is an element of the tree system, which is physical. Photosynthesis in the tree is part of the system as well, but you cannot see or touch it, it is amorphous.

All elements of a system are connected through relationships. In the example of a tree, these relationships might be chemical reactions or other processes like photosynthesis.

If we want to learn the purpose of a system, we need to observe its behavior.

A system is always defined by the purpose of the system and the relationships between elements in the system.

Stocks and flows which change over time define the system's behavior.

Again let's look at an example to make this clearer. In the system of a book store, the books standing on the shelves is the stock. The flow is the change of stock, or in this example, the inflow and outflow of books. When a customer buys three books, the inventory will decrease. The stock increases when the owner of the book store orders new books.

Balancing and reinforcing feedback

When the stock of a system changes through flow (inflow or outflow), the system receives feedback.

There are different kinds of feedback. There is, for example balancing feedback and reinforcing feedback.

Feedback is known as balancing feedback when it stabilizes the stock of a system. There is a relation between the stock and the input which regulates the flow. This relation can be physical laws or rules.

Reinforcing feedback changes the stock of a system. The stock can increase perpetually by the feedback, or it can continuously decrease the stock until it is used up and the system collapses. For example, when you accrue interest for your money in your bank account, your balance will increase. And the more money you have in the account, the more interest you earn on your money.

Often systems have both balancing and reinforcing feedback.

Well-functioning systems

When a system can adapt to changing conditions, it is resilient. This resilience is a primary characteristic of well-functioning systems. This resilience determines how well a system responds to changes.

If we want a system to keep running, we should pay close attention to the system's resilience. Often we sacrifice resilience to other goals. An example might be that we often sacrifice our sleep to work more. But the work's output will suffer if we don't get enough recovery time. And if we deprive our body of sleep for a too long period, it might even collapse.

Another essential characteristic of a well-functioning system is self-organization. Complex structures are built of self-organized subsystems that are hierarchical. The hierarchy is functional because it reduces the amount of information that has to flow between subsystems.

Common mistakes when assessing systems

To better understand systems, we need to watch how it's functioning over time. We have to watch the behavior of the system. Sometimes we oversimplify things as we only assess the outcome of a system but ignore the real behavior.

Another mistake we often make when evaluating a system is that we think that relationships are linear. Assuming that all systems are linear is a mistake because most things in the world are not.

A third error we make assessing a system is that we mentally isolate a system. But most systems are not separated from other systems. We oversimplify things resulting in wrong conclusions.

How corrupt systems are produced

Sometimes a well-functioning system is corrupted by one of its subsystems. This corruption happens when a subsystem gets to much power and uses it to force the whole system in a new direction. When this happens, all other systems have to work harder to keep the system's original behavior. The result will be a system that is stuck and keeps producing unwanted outcomes.

It is necessary to re-organize all involved systems and use the energy to build new relationships between the subsystems and elements if we want to fix a corrupted system.

Another cause why systems fail is that it is using unsustainable resources. The system will function well until the supply is drained. This failure might happen because of a lack of knowledge or ignorance of the involved actors.

Improving systems

We can improve systems and make them produce more of what we desire and less of what we don't want.

To make a system better, we can change delays, buffers, and system design.

Every system has delays. These delays can make it hard for a system to respond to changes at the proper time. Say you have a system with long term delays. For such a system, it will be a challenge to respond to short-term changes. Therefore delays can be an adjusting screw to make a system more effective.

Time, inventory, and storage space are system buffers. For a system to work at the highest efficiency, we must ensure that these buffers are correctly sized.

System design is yet another possibility to improve a system. A well-designed system knows its limitations and is less prone to fluctuations.

We can improve systems even further by adjusting its rules, information flow, and self-organization.

Sometimes a system just needs more information to get better. If we recognize which information is missing and make it available to the system, it can improve itself.

Some rules restrict the self-organization of a system. Therefore a system can organize itself better when these rules get removed.

Learn to understand systems

We have an urge to learn and understand, but we can comprehend systems only in a general sense. We want to be in control, but we cannot fully control systems. Still, we can get this broad understanding of a particular system and improve it.

There are simple steps to enhance your understanding of systems and even find ways to improve them.

collect data about a system observe a system and understand its history make models of a system you want to understand pay attention to what is vital for a system to function pay attention to things that can be measured and things that can't be measured that easily, like quality and justice for example

In addition to the listed points to understand a system, you need to focus on things that influence the behavior of a system and ask yourself if these influences are controllable.

First Insight: We cannot fully understand systems

A system must consist of three kinds of things: elements, interconnections, and a function or purpose.―Donella H. Meadows

Systems consist of elements, relationships between those elements, and a purpose that is defined by its behavior. This might be the reason that we think we can fully understand a system. We even go one step further and believe that we can control a system. But we can't.

The only thing we can do is to improve a system or break it.

Second Insight: A systems' purpose is deduced from its behavior

Purposes are deduced from behavior, not from rhetoric or stated goals.―Donella H. Meadows

It is vital to understand that actors within a system might proclaim that they will do X, but when the system's behavior does not support it, X will never be done.

It makes no difference if an actor within a system makes a rhetoric statement. If you want the system to do X, you need to change the elements and the relationships within it to change the system's behavior.

Often this is not quickly done because we seldom fully understand and control a system.

Third Insight: To understand a system we need to make a model

Remember, always, that everything you know, and everything everyone knows, is only a model. Get your model out there where it can be viewed. Invite others to challenge your assumptions and add their own.―Donella H. Meadows

To understand a system, we need to come up with a model. But we cannot expect to get this model right without the feedback from other observers of a system.

Next time I try to understand a system, I will make a model and show it to the people around me.


Systems are everywhere. Most of the things we experience or do are the result of systems in action. When we try to understand a system, we probably will not understand it and its relationships. Still, trying to understand systems is critical to make adjustments, repair broken systems, and follow the world better in which we live.


  • To keep a system running, do not sacrifice its resilience to other competing goals
  • Do not mentally separate connected systems while learning about a system
  • Try to understand systems around you
  • Support self-organization

Thinking in Systems Summary—related resources

  Read more about Donella H. Meadows on Wikipedia.

If you liked reading the Thinking in Systems Summary, you should also read Thinking, Fast and Slow Summary.

Posted on CuteMachine in nature and science.

Jo's Profile ImageWritten by Jo who lives and works in Frankfurt building digital doodah. Stalk him on Twitter.

TOS • Privacy • Disclaimer