Chapter 2: Systems Thinking

Tai Munro

Key Ideas

In this chapter, you will learn about:

  • wicked problems
  • systems thinking
  • parts of a system

Wicked Problems

Sustainability problems are what are known as wicked problems. There are formal definitions of wicked problems (e.g., Rittel & Weber, 1973), but generally speaking, they are problems that:

  • require many different people to be involved
  • we have incomplete information for
  • the requirements are contradictory or changing
  • there is no single solution for
  • require a culture shift to address

Most problems that we define as wicked, like climate change, addiction, and inequity, have ways of thinking and knowing that influence them. For example, we might agree that we need a program to support people suffering from addiction, but factors like whether we believe that addiction is caused by differences in brain function or by lack of learned impulse control will impact how we design addiction support programs.

These characteristics make wicked problems resistant to being resolved. Let’s take two problems as examples.

Image by congerdesign. Shared under a Pixabay License.

Not Wicked: Difficulty Getting Up in the Morning

While to the individual, this is a pretty challenging problem, in relation to problems, this one is pretty easy to solve. People can easily collect data on things like when they go to bed, how long it takes them to get to sleep, and how many times they wake up at night. Then, without consulting anyone else, they can try some different solutions like going to bed early, starting or stopping a bedtime snack, wearing headphones while they sleep or using a white noise generator app, and trying a wake-up light.

They can experiment until they find success and then hopefully continue the habits that created that success.

Image by Agnieszka. Shared under a Pixabay license.

Wicked: Ending Homelessness in Your City

Different types of homelessness need to be addressed (chronic and episodic); there are many different stakeholders and partners; homelessness can be connected to the availability of affordable housing, unemployment, support for physical and mental health concerns, etc.

How do we experiment with the different factors? How do we know how much we need to change each element? How do we know that we have ended homelessness?

Western Approaches to Problem Solving

In Western problem-solving traditions, we tend to break things down, or reduce them, to their parts and then look at them individually. Capra and Luisi (2014) describe this as a focus on “What is it made of?” or the material objects and structures (p. 4). This approach works well for problems like not being able to get up in the morning. But, by ignoring the interconnections between the different parts or what Capra and Luisi (2014) describe as the non-material processes and patterns of organization, it is generally insufficient for dealing with wicked or complex problems like ending homelessness.

Some of the consequences of this traditional approach to problem-solving include:

  • We focus on finding answers or solutions (even if we don’t fully understand the problem) (Ackoff, 1978).
  • We assume that it is easy to trace a problem to its cause (Stroh, 2015).
  • We are often unaware of the systemic structures that influence behaviour over time. We “just find ourselves compelled to act in certain ways” (Senge, 2006, p. 44).
  • We believe that we can optimize the whole by optimizing the parts (Stroh, 2015).
  • We ignore the context of the larger whole (Capra & Luisi, 2014).

To overly simplify this challenge, traditional problem-solving approaches ignore the context and relationships that influence what and how problems occur. Systems thinking is a way of addressing this challenge.

It should be noted that while this approach, often referred to as reductionism or a mechanistic view, has been dominant in Western culture for the past 300 years, since the Scientific Revolution, it had previously competed for prominence with a more holistic view (Capra & Luisi, 2014). In addition, thinking in many other cultures has not focused so heavily on this mechanistic approach (Capra & Luisi, 2014).

Introducing Systems

Before we look at systems thinking, we have to start by asking what a system is. Meadows (2008) defined a system as “an interconnected set of elements that is coherently organized in a way that achieves something” (p. 11). This is a good place to start. It tells us that a system includes elements or parts that are connected in such a way that they achieve something. The interconnections or relationships are important to the system. And the system is going to achieve something. But, a few key characteristics aren’t clearly captured in this definition.

Many of us understand the word “organized” as meaning that an actor organizes something. However, a system is self-organizing. This means that it is dictated or organized by internal rules rather than external forces (Capra & Luisi, 2014). Thus, the idea of organization in a system can be confusing. While a system, such as a traffic system, might include organizing elements that have been planned by humans, like traffic laws, the system as a whole is governed by self-organizing features that arise from all of the interconnections in the system. This is one reason why certain roads tend to have higher rates of speeding even though the traffic laws penalize speeding.

Another challenge in Meadows’ definition is the idea that the system “achieves something.” Similar to “organized,” the problem isn’t inherent in the words but in how we understand them. “Achieves something” guides us to believe that a system will achieve a particular goal and that that goal will be what we want it to be. However, this is not the case. Many systems do not achieve something desirable. Homelessness, addiction, poverty, injustice, environmental degradation, and climate change are all achievements of human systems that we do not want. Meadows (2008) discusses this characteristic of systems, but the challenge with our common understanding of the language remains.

In their efforts to trace the history of the development of systems thinking, Capra and Luisi (2014) define a system as “an integrated whole whose essential properties arise from the relationships between its parts” (p. 64). This shares a lot with Meadows’ definition but avoids some of the challenges embedded in our understanding of the language. At the same time, it is helpful to consider both definitions, as they reveal more about what a system is. A system:

is made up of elements

that are interconnected with each other

to create an integrated whole

that, through self-organizing and emergence,

result in properties that are not found in the individual elements.

Let’s spend a moment looking at what self-organizing and having emergence mean.

Self-organization is a property where the organization of the whole arises from the interactions of the parts of the system. You may also hear this referred to as autopoiesis which refers to the ability of something to reproduce itself. But reproduction here should be clarified. It is not reproduction in the sense of making another organism, system, or offspring of any sort but of its ability to maintain itself over time. An example may help clarify here. A sports team regularly changes individual players, but this does not often result in a significant change to the team. This is because the goals and behaviours of the team are governed by the interactions in the system, and individual players rarely impact these in significant ways. Or, as Pirsig wrote, “if a factory is torn down but the rationality that produced it is left standing, then that rationality will simply produce another factory” (as cited in Meadows, 2008, p. IX).

Emergence occurs when novel system properties arise “from the specific relationships and interactions among the parts in the organized ensemble” (Capra & Luisi, 2014, p. 155). In other words, the system’s properties cannot be isolated to any of the individual parts. While emergent properties can be what makes something unique such as how your genetics, physiology, upbringing, education, and other factors, come together to make you a talented artist, athlete, leader, or another role, emergent properties can also be the properties that we don’t want like homelessness and environmental degradation.

Systems Thinking

Einstein is famously quoted as saying, “we cannot solve our problems with the same thinking we used when we created them.” As different fields have developed, many have sought out new ways of thinking to solve difficult problems. Accordingly, the history and practice of systems thinking include many fields, such as quantum physics, cybernetics, ecology, and management and leadership (Capra & Luisi, 2014; Senge, 2006). The challenges of sustainability, from climate change to food security, grew out of the mechanistic approach to thinking. Therefore, we must adopt a new way of thinking to solve these challenges. Systems thinking is most commonly identified as an appropriate approach (Evans, 2019; Redman & Wiek, 2021). Or as Capra and Luisi (2014) put it, “the systemic understanding of life… is the cognitive foundation of our endeavor to move towards a sustainable future” (p. 362). But what exactly is systems thinking?

Let’s start with a few definitions again:

  • “The understanding of a phenomenon within the context of a larger whole” (Capra & Luisi, 2014, p. 64)
  • Building on the definition of systems from Meadows, Stroh (2015) defines systems thinking as “the ability to understand [the interconnections between elements] in such a way as to achieve a desired purpose (p. 16, italics original)
  • “Systems thinking is a conceptual framework, a body of knowledge and tools… to make the full patterns clearer, and to help us see how to change them effectively” (Senge, 2006, p. 6)

What do you notice between these three definitions? What do they share? What is different?

Like the second and third definitions, Russel Ackoff also includes the ability and intention to change systems as part of his definition. Watch the following video (4:05) to learn more about the differences between reductionism and systems thinking and how Ackoff wants us to think about changing systems.

Reflection 2.1: Solving Problems

Can you think of a time when you tried to solve a problem that seemed resistant to change? You knew there was an issue, but you couldn’t do anything effective to change it? What strategies did you try to solve the problem? Do you think these strategies demonstrate reductionist or systems thinking?

While a definition is helpful, it doesn’t tell us much about how we might do systems thinking. And indeed, the descriptions of how to do systems thinking are almost as numerous as the fields that engage in systems thinking. However, there are some common characteristics.

Thinks about the whole: Not surprisingly, considering the focus on the interconnectedness of systems, a common characteristic of systems thinking is that we need to shift our perspective from the individual parts to the integrated whole. This includes recognizing that humans and individuals are parts of the system (Sweeney & Meadows, 2010). This is not to say that we do not need to think about the parts but that we need to consider the interplay and relationships between the parts and the whole (Capra & Luisi, 2014).

Integrates multiple perspectives: Team learning (Senge, 2006) or multidisciplinarity (Capra & Luisi, 2014) are needed to move beyond individual perspectives or disciplines.

Maps relationships: While we may be able to quantify individual items, relationships need to be examined qualitatively. Further, we need to look for and identify “interrelationships rather than linear cause-effect chains” (Senge, 2006, p. 73). To support pattern identification and understanding, we must map the system’s relationships and causal loops (Senge, 2006; Sweeney & Meadows, 2010; Capra & Luisi, 2014).

Examines mental models: We all have “deeply ingrained assumptions, generalizations, or even pictures or images that influence how we understand the world and how we take action” (Senge, 2006, p. 8). In systems thinking, we must consider how these mental models create our futures (Sweeney & Meadows, 2010). By examining mental models, we can then challenge them. Mental models show up in many ways, such as stereotypes.

Considers the long-term: We need to acknowledge that short-term solutions may not be solutions for the long term (Sweeney & Meadows, 2010). As Senge (2006) states, “vision without systems thinking ends up painting lovely pictures of the future with no deep understanding of the forces that must be mastered to move from here to there” (p. 12). This also includes paying attention to time delays that occur within systems and can hide consequences over the short-term (Meadows, 2008; Sweeney & Meadows, 2010). We need to look for “processes of change rather than snapshots” (Senge, 2006, p. 73).

The Parts of a System

To help us talk about and map systems, as we will do in the next chapter, it is helpful to establish some common language. Different sources use different terms to describe the same things. Therefore, it is also important to pay attention to the general characteristics included in each term so that you might be able to transfer what you are learning to other resources and materials in the future.

Image by Tai Munro


The elements of a system are often tangible things like the paddlers in a boat, the boat they are paddling, the water body they are on, and the weather they are experiencing. However, elements, such as the paddlers’ skills, can also be intangible.


The interconnections are “the relationships that hold the elements together” (Meadows, 2008, p. 13). In the boat example, the paddles apply force to the water that moves the boat because the paddlers are in the boat; they are interconnected. The paddlers have to communicate with each other. The wind may push against the boat, affecting its path or how difficult it is to move. As Meadows (2008) points out, you can change the elements without drastically affecting the overall system; however, the interconnections impact the system’s emergent properties and, therefore, tend to have a more significant impact on the system’s outcomes.

See if you can answer the following questions about elements and interconnections.

Emergent Properties

Emergent properties were discussed earlier in this chapter. Emergent properties are properties that arise from a system but are not properties of individual elements or interconnections. Systems thinkers may assess emergent properties as positive, negative, or neutral.

For example, presumably, the goal of a food bank is to alleviate food insecurity by providing access to food to people in need. However, we must ask whether a food bank changes how often or why people experience food insecurity. A food bank can only really provide a temporary solution to the existing problem in the system. That existing problem is an emergent property of the system. Without change, the system will continue to produce conditions that result in economic inequality and food insecurity.

Food Banks Canada has recognized this. One of their major goals is to find “long-term policy solutions to food insecurity” in Canada (Food Banks Canada, 2022, Policy and Advocacy, para. 1). This is an attempt to change the system so that food insecurity is no longer an emergent property.

Feedback Loops

Feedback is one of the key parts of systems. However, feedback is challenging to see because we are used to seeing causality as a linear process (Senge, 2006). Let’s break down a common experience in school to help us understand feedback loops.

A linear process showing Student submits Assignment which informs the Professor who provides Feedback

As presented, we see a linear process. A leads to B which leads to C. When we get to the end, the professor provides feedback, but the causal chain stops there. Unfortunately, although you probably don’t do this, an all too common occurrence is that the student checks their grade on the assignment and then doesn’t do anything else with it. They might not even read the feedback. But feedback in systems thinking is “any reciprocal flow of influence” (Senge, 2006, pp. 74-75). In other words, feedback is something that happens in such a way that both the initial element and the subsequent elements influence each other in related ways. So let’s look at what this would look like to a systems thinker.

A circular process showing student submits assignment to professor gives feedback to student so that they cycle continues

What wasn’t apparent when we looked at the process in a linear diagram is that the feedback should influence the student. This is the idea of using the feedback that you get on one assignment to either improve that assignment or to help you with your next assignment.

Watch the next video (1:20) to learn more about feedback loops.

Note: The following video is a whiteboard-animated video. The illustrations used depict what is being described in the narration. You do not need to see the illustrations to understand the video.

It isn’t surprising if you had trouble seeing the loop part of causal loops. How we are taught, and the English language structure both teach us to see causality as a linear process rather than a circular one (Senge, 2006). In addition, we typically see humans as separate from systems and the feedback process in Western culture. Shifting our awareness so that “the human actor is part of the feedback process… represents a profound shift” (Senge, 2006, p. 77, italics original) and one that can be difficult to make.

We will look more at feedback diagrams in the next chapter. For now, let’s move on to looking at the two main types of feedback that exist in systems.

Balancing Feedback Loops

A balancing feedback loop occurs when there is a goal or target for the level of an element, and the changes occur in such a way that the level always stays around that goal. There are many examples of balancing feedback loops, such as:

  • your eyes adjusting how much light they let in based on the brightness of the space you are in
  • eating when you are hungry
  • keeping a certain amount of stock of different products on a shelf
  • the number of a single species in a habitat
  • the temperature of your cup of coffee

An important point regarding balancing feedback loops is that “the system has its own agenda” (Senge, 2006, p. 84). Thanks to the laws of thermodynamics, that cup of coffee you have sitting with you while you read this has a goal of being at room temperature. Unfortunately, this means that the real goals of the balancing process are often hidden. Watch the next video (2:37) to learn more about balancing feedback loops.

Note: The following video is a whiteboard-animated video. The illustrations used depict what is being described in the narration. You do not need to see the illustrations to understand the video.

One reason balancing loops are hard to identify is that it often “looks like nothing is happening” (Senge, 2006, p. 86, italics original). Unless you have diabetes or another condition that impacts your blood sugar levels, you probably don’t really notice any changes in your blood sugar throughout the day. This doesn’t mean that it isn’t changing. But your body maintains the balance well enough that you don’t notice the small changes that are happening. Similarly, the sales of a particular product tend to stay at the status quo, making it hard to see what factors might be impacting product sales.

Reinforcing Feedback Loops

A reinforcing feedback loop occurs when there are runaway effects (exponential growth or collapse). If changes in an element result in more changes in the same direction, you get a reinforcing feedback loop. In other words, a small change will build on itself in an accelerating manner. You might have heard of these in other terms, such as self-fulfilling prophecies or vicious circles. An example of this is when something goes viral: people like something, which results in more people seeing it, which results in more people liking it, which results in more people seeing it, and so on.

A big challenge with reinforcing feedback is that, up to a point, the changes are small enough that they are hard to detect, or they just don’t seem like a big deal. Watch the next video (1:33) to learn more about reinforcing feedback loops and review some examples.

Note: The following video is a whiteboard-animated video. The illustrations used depict what is being described in the narration. You do not need to see the illustrations to understand the video.

See if you can answer the following questions about feedback loops.

Causation Versus Correlation

You may have heard about the difference between causation and correlation before. Correlation means that there is a relationship between two things, but one does not cause the other. Causation means that one thing causes something else to happen. Correlation often means that another element is connected to the two things you are looking at. The difficulty in distinguishing between these two can make it harder to find the actual feedback loops in a system because you need to find causation, not correlation. Watch the video (2:33) to learn more.

Note: The following video is a whiteboard-animated video. The illustrations used depict what is being described in the narration. You do not need to see the illustrations to understand the video.

Leverage Points

Leverage points are points where you can push a system in order to trigger systems change. You probably use leverage points all the time without thinking about it. Let’s consider some examples of leverage points in different situations.

  • A leverage point that students often try to use is to study for more hours. They aren’t happy with their grades, so they dedicate more time to studying. This can have an impact, but there might be better leverage points to push, such as studying for the same amount of time but using more effective study strategies or meeting with their professor and asking specific questions. (In case you are interested in what might be some more effective study strategies, you can check out the Learning Scientists videos.)
  • One leverage point that sports teams often attempt to use is bringing in a star player to revitalize the team. This, in theory, is supposed to make the whole team better.
  • Applying carbon taxes or funding solar panels or energy-efficient appliances are leverage points intended to incentivize better practices or equipment.
  • Systems thinking itself is a leverage point. By looking at challenges through a systems lens, we can identify new ways to intervene in a system. For an example of this in health care, watch the video Systems Thinking! posted by james swanson.
  • Another leverage point is to use other ways of looking at challenges, such as Indigenous perspectives, disability perspectives, or gender equity perspectives. The town Karlskoga in Sweden used a process of gender mainstreaming to assess snow removal practices in the town. The town’s snow clearing policies prioritized clearing snow off main roads after a snowfall. Because women in Karlskoga are more likely to travel by foot, public transit, or bike, and men are more likely to travel by car, the snow-clearing policies affected men and women differently. An emergent property of the system was prioritizing men’s daily activities and safety over women. By using gender equity as a leverage point, the town changed their snow removal practices to prioritize sidewalks and residential areas over main roads. As a result, hospital costs in Karlskoga declined because there were fewer injuries, primarily to women, from walking on local icy sidewalks and no real increase in car accidents on the main roads. You can watch the first 3:17 minutes of the video Sustainable Gender Equality – a film about gender mainstreaming in practice published by SKR Jämställdhet to learn more.

You may have noticed in reviewing these examples and in thinking of some of your own that while there are many potential leverage points, they don’t all have the same level of impact. Senge (2006) suggests that “our non-systemic ways of thinking consistently lead us to focus on low-leverage changes. Because we don’t see the structures underlying our actions, we focus on symptoms where the stress is greatest” (p. 113). A good example of this is food banks, which were discussed earlier in the chapter. A food bank is a leverage point to alleviate food insecurity on a short-term or temporary basis. It addresses a symptom of the problem so that fewer people go hungry, but it does not address the reasons that food insecurity occurs in the first place.

Another example of low-leverage versus high-leverage changes comes from agriculture. In the short term, adding synthetic fertilizers can help increase yield, but it does not address the systemic issues causing declines in soil health. In fact, synthetic fertilizers can contribute to declines in soil health. This means that although applying these fertilizers is a short-term leverage point to increase yield, it also creates a reinforcing feedback loop that requires more and more fertilizers and other additions to maintain yields over the long term.


Before we move onto systems mapping, Kalen Pilkington’s TedX Talk (11:08) from the MacEwan TedX event in 2018 provides a great review of systems thinking and a concrete example from her undergraduate student experience. As you watch, try to identify the part of the system we just reviewed: elements, interconnections, emergent properties, feedback loops, and leverage points.

See how you did with identifying the different parts of the system.

Reflection 2.2: Leverage Points

Think about an area in your life where you would like to achieve change. See if you can identify three possible leverage points that you could apply a change and trigger a significant impact on the overall system. Some examples that you might want to consider include making more time for friends and family, improving your overall fitness and health, or improving your performance at school or work.


Capra, F. & Luisi, P. L. (2014). The systems view of life: A unifying vision. Cambridge University Press.

Evans. (2019). Competencies and Pedagogies for Sustainability Education: A Roadmap for Sustainability Studies Program Development in Colleges and Universities. Sustainability, 11(19), Article 19.

MacEwan University. (2018). Systems Thinking [Video].

Meadows, D. (2008).  Thinking in systems: A Primer. (D. Wright, Ed.). Chelsea Green Publishing.

Redman, A. & Wiek, A. (2021). Competencies for advancing transformations towards sustainability. Frontiers in Education, 6: 785163.

Rittel, H. W. J., & Webber, M. M. (1973). Dilemmas in a General Theory of Planning. Policy Sciences, 4(2), 155–169.

Senge, P. M. (2006). The fifth discipline: The art & practice of the learning organization (Rev. ed.). Currency.

Stroh, D. P. (2015). Systems thinking for social change: A practical guide to solving complex problems, avoiding unintended consequences, and achieving lasting results. Chelsea Green Publishing.

Sweeney, L. B. & Meadows, D. (2010). The systems thinking playbook: Exercises to stretch and build learning and systems thinking capabilities. Chelsea Green Publishing.

TEDx Talks. (2018, February 15). A Systems Thinking Approach to Community-Based Urban Agriculture | Kalen Pilkington [Video]. YouTube



Icon for the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License

Introduction to Sustainability Copyright © 2023 by Tai Munro is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License, except where otherwise noted.

Share This Book