I believe it is important to teach oneself and others how to think as opposed to what to think. Instead of thinking in parts, focusing on the interactions and relationships help prime the mind to identify patterns more easily, and thus build stronger capabilities for inference. The mind becomes critical and creative.
I’m not the only person to think this; educational frameworks have emphasised the importance of systems thinking in recent years. In this post, I highlight some of the personal approaches I used in my attempts to teach systems thinking.
Understanding Systems #
It is imperative that to teach systems thinking, one must be a systems thinker.
The basis of a system consists of its elements, relationships between the elements, and an overall function of the system. Thinking in Systems: A Primer by Donella H. Meadows expands on this and says that a system is composed of stock(s) and its flows. A stock is a measurable element of a system, while flows such as inflows and outflows are factors influencing the stock.
block
inflow<["inflow"]>(right) stock outflow<["outflow"]>(right)
Systems can have more than one stock, and more than one inflow or outflow. An example to make this easier would be population. You have a stock of people occupying a geographical region, and their numbers could be added through births. Births represent the inflow. Similarly, deaths could reduce the population, thus representing the outflow. Both of these control the stock. Migration into the population could be another inflow, while migration towards outside the population is another outflow.
The stock typically changes gradually. You could have somewhat similar birth rates and death rates, leading to a dynamic equilibrium in which the stock retains itself. However, it is entirely possible that a disaster could lead to a sudden increase in death rates, thus decoupling the outflow and resulting in a drastic change in stock.
Inflow and outflow are influenced by feedback, and two types of feedback occur: positive and negative feedback, also known as reinforcing and balancing feedback loops respectively. A positive/reinforcing feedback loop typically enforces a certain behaviour. For example, more people would result in more children and more children will result in more people. A negative/balancing feedback loop generally suppresses a certain behaviour. For example, having too much coffee might cause insomnia, and this informs your choice to have less coffee next time.
Now that we have understood the concept of systems, we will build from ground up, this model of stocks, inflows, outflows and feedback.
Level I: Identity of Stock #
At a conceptual level one must learn to identify the “stock”. This is incredibly simple: Nouns generally describe stocks, and this can be taught contextually through language studies and grammar.
Stocks must be able to be differentiated from other stocks: Describe an entity and compare it by describing a different entity. Specific combinations of descriptors could be used to define identity of the stock.
Introduction of the “opposition” and “negation” of a stock is not necessary, but recommended.
- A stock can be represented by a positive number 1.
- Its opposition is the negative number -1.
- Its negation is 0.
Differentiate between negation and opposition: the negation of red could be colourless while its opposition is blue.
Level II: Relationships #
You would want to introduce the conceptual understanding of relationships, and with it you introduce two of the components of a system: its elements and its interconnections. This can vary in complexity. A simple if-then statement could be a starter point:
flowchart LR
id1[it rains] -->|causes|id2[open an umbrella]
This shows a causal relationship with two elements. If-then statements are great for building the skill of inference, but you should not be limited by if-then statements.
Introduce a non-causal relationship such as parentage and distinguish it as distinct.
flowchart LR
Mark -->|father of|Timmy
Further along you would want to introduce a chain of statements to introduce systems with more than two elements. This helps introduce sequential logic in a systems thinking perspective.
Using if-then statements are beneficial because you could use if-then statements to also introduce sequential logic or the law of logical consequence. The law of logical consequence states that provided a set of premises is true, a sentence is either a logical consequence (or not). So for example, if it’s raining and you have an umbrella open, you’re either dry or you’re not, in which being dry is the logical consequence.
flowchart LR
id1[it rains] -->|causes|id2[open an umbrella]
id2 -->|causes|id3[remain dry]
The if-then chain is congruent to a stock-flow diagram in which the inflow is the amount of rain weathered by a person and the outflow is the presence of umbrellas, with the stock being the degree of wetness the person will have. While the previous model is sufficient in introducing relationships, stock-flow models could be beneficial for introducing more complex examples such as population of a country or flow of economy.
block
rain<["rain"]>(right) wetness umbrellas<["umbrellas"]>(right)
Level III: Feedback #
Systems could reinforce themselves or self-regulate. Reinforcing feedback loops reinforce, balancing feedback loops regulate. At this stage, practice by analyzing and creating causal loop diagrams (CLD) on common phenomena to grasp the concept of feedback. Mastering this method will help build the ability to visualise CLDs mentally.
Start with a simple description.
When there are a lot of births, the population increases, which leads to increased births. However, when there is a high population, there would be more deaths, which would decrease the population.
Break down the text into statements with components, highlighting relationships such as increase or decrease.
- Births increase populations.
- Populations increase births.
- Populations increase deaths.
- Deaths decrease populations.
You can begin building the CLD after you have identified the components and relationships. It is important to introduce the rules of CLDs at this stage. Refer to Guidelines for Drawing Causal Loop Diagrams.
flowchart LR
births -->|+| population
population -->|+| deaths
deaths -->|-| population
population -->|+| births
Another method is to take a complex CLD and analyze it. For example, use the UK Government’s CLD on diabetes. For this, do the reverse: analyze the CLD, identify the statements, and then build a description.
Interchanging sequential logic and systems thinking #
A complex CLD is also capable of representing a chain of sequential logic; for example, some parts of the diabetes CLD from the UK government could be defined with these statements:
- If a person lacks education, their perceived value of activity could be low.
- Low sociocultural valuations of activity can lead to low levels of recreational activity.
- Low levels of recreational activity can decrease physical activity.
- Functional fitness decreases with physical activity.
- Low physical activity can lower self esteem.
- And so on, so forth.
Level IV: Perspective #
In a classroom setting, it is feasible to obtain various samples of CLDs drawn by students and spotting the differences between different student work. This introduces the concept of Perspective.
A system could be likened to a 3D structure which, when viewed from different view points, projects and entirely different 2D image. Mental reconstructions of the system provide the 3D element, however, there is a process of flattening the system that provides the element of perspective.
Equivalence to DSRP method #
The DSRP method is a framework by systems scientist Derek Cabrera and is one of the most famous and well-tested frameworks available. DSRP stands for Distinctions, Systems, Relationships and Perspectives which composed the 4 patterns of the DSRP and each pattern has two paired elements.
- Distinction: identity co-implies other
- Systems: parts co-implies whole
- Relationship: action co-implies reaction
- Perspective: point co-implies view
This is directly represented by the levels in this article: Distinction is represented in Levels I and IV, Systems and Relationships are represented by Levels II and III, while Perspectives is represented in Level IV.
Conclusion #
Mental scaffolds are important in order to introduce, gradually, systems thinking to a general audience. An approach with levels would help foster a conceptual understanding of systems thinking.
While the approach I have shown is only one of multiple approaches, I hope this manual may serve as guidance for implementation of other methods.