"Dana" Meadows Provides a Primary Systems Thinking Review pt 3

It is about time for the usual reminder given with posts on Systems Practice and other courses. This blog post series on Donella Meadows’ book “Thinking in Systems - A Primer” is not a substitute for reading the book. Read the book. This is only my perspective on what I have learned from Donella Meadows’ lessons in particular and Systems Thinking in general.

The last post discussed what was important to discern about systems (purpose or function), the constrained roles of leaders, and feedback loops, particularly balancing loops used to regulate. Balancing feedback loops though don’t work only through human or programmed decisions, they also work through physical laws.

“Whatever the initial value of the system stock (coffee temperature in this case), whether it is above or below the “goal” (room temperature), the feedback loop brings it toward the goal. The change is faster at first, and then slower, as the discrepancy between the stock and the goal decreases”.

It is based on a function as discussed in the blog posts on the Complexity Explorer course.

We are used to (I am) thinking of goals as being tied to a purpose (the purpose of this activity is to attain that goal), so something is imposed to change the current situation from what it is to what we want. Which may be part of the reason we often have difficulty discerning purpose or function, if it’s a matter of maintaining the system’s own integrity rather than what we want.

The goal or better function (the purpose is the goal) of a river is to flow into a lake or ocean. The river does not do this intentionally on purpose. Rather it is a step or function in a feedback loop of the world’s hydrology system and subsequently an element in many other systems, including ecological, fishing and shipping.

Meadows makes another important point, That balancing feedback mechanisms don’t always work, at least not as we want. They may not be capable of bringing the stock to the desired level, interconnections, especially the information part of the system may fail, arrive too late or at the wrong place or be unclear or incomplete or hard to interpret. Actions triggered may be too weak or delayed or be resource-constrained or simply ineffective. Breakdowns in a system should be differentiated though from poor system design.

Goal-seeking or stabilizing balancing loops are not the only type of feedback loops. Reinforcing loops, the second type of the two, exist when an element (or stock) within a system has the ability to reproduce itself growing at a constant fraction of itself as with populations and economies.

They are amplifying, reinforcing, self-multiplying. They can either cause healthy growth in a virtuous circle or runaway destruction in a vicious one enhancing the direction of change that is imposed upon on it with the capacity of snowballing. Generating more input to a stock the more it has and less input the less it has with each iteration.

This is not simple linear growth at a constant rate over time. It is exponential growth. Again, following the path of functions like the logistic equation discussed in the Complexity Explorer series.

I believe that another point can be derived from what Meadows has said about feedback loops and the connections or flows through which they operate. They invariably require energy. Reinforcing loops always require energy, as with life sustained through energy from the sun for populations. A balancing loop may simply be turned on or off based on attaining a goal (information feedback) but often can also require additional energy to act as a countervailing force.

Think of Meadows’ example of a coffee cup cooling but replace it with an unplugged refrigerator being defrosted. The function remains the same. The change from the initial value to goal follows the same functional path. It can also be considered a balancing loop. Plugging the refrigerator back in involves applying energy. Electric energy is converted to mechanical energy which is used to cool again the food inside. A decision is made as to where to set the desired temperature and another balancing loop is put into effect.

The first balancing loop used the second law of thermodynamics allowing the refrigerator to go naturally from a lower state of entropy to a higher one. The second balancing loop, still following the second law, forced the closed system through the input of energy into a lower state of entropy although it increased overall (higher) entropy in the environment.

Reinforcing loops in this scenario could be the increasing use of energy to power the refrigerators because of increasing population and decreasing resources. More people means more refrigerators bought especially if economic circumstances improve for some and that means more energy consumed depended upon dwindling fuel supplies. Real world, physical systems will involve entropy. This could be considered a source for so-called unintended consequences of systems.

A stock then can have several reinforcing and balancing loops of differing strengths pulling it in several directions. A system’s different feedback loops can make the systems stocks grow, decrease until eliminated or coming into balance with other flows in the system. A flow may be adjusted by the contents of a number of different stocks, filling one stock while draining another while feeding (via information) into decisions affecting yet another one.

Feedback loops are easier to understand as graphics such as systems maps than as written words. ”Pictures work for this language better than words, because you can see all the parts of a picture at once”. This means though learning, as was said in the first post of this series, a new way of doing things, a new way of thinking, a new cognitive grammar capable of inquiring into and understanding systems. There are online programs and communities attached to them that can help with this. My go-to choice is Kumu.

However, anyone somewhat familiar with this New Community Paradigms effort in general and more specifically with the approach to Systems Thinking and of Systems Practice will know that neither the Acumen course nor usually I, use Stock and Flow models. (When I do, I use Insight Maker) The Systems Practice course and I instead use Kumu to create Causal Loop Diagrams (CLDs).

CLDs include (conceptually) elements and connections as do Stock and Flow Models but they don’t have explicit quantifiable stocks, and while flows can be quantified, connections in CLDs are not. The preference for CLDs is that they are conceptually easier to work with and more intuitive for others to understand. However, Meadows (and others) has convinced me that we should at least be thinking of what the stocks would be, whether physical or intangible and what paths the connected flows would take to ascertain that the system in question is having a manifested and not an imaginary impact on the world.

Meadow speaks of challenging her students (and readers) to think of human decisions that occur without a feedback loop, a decision made without regard to information about the level of the stock being influenced with “falling in love” and “committing suicide” being the most common supposed “non-feedback” decisions. The motivations for both are intangible so we can have difficulty with thinking in terms of stocks and flows but our world is materially different when the state of the world exists either with them or without them and if it does with either that it can quickly snowball in that particular direction upon reaching a tipping point.

In real systems, feedback flows are linked together, often in fantastically complex patterns going beyond being limited to only one of three basic states of being steady or approaching goals smoothly or exploding exponentially.

”Watch out! If you see feedback loops everywhere, you’re already in danger of becoming a systems thinker! Instead of seeing only how A causes B, you’ll begin to wonder how B may also influence A—and how A might reinforce or reverse itself”. When you hear in the nightly news that the Federal Reserve Bank has done something to control the economy, you’ll also see that the economy must have done something to affect the Federal Reserve Bank. When someone tells you that population growth causes poverty, you’ll ask yourself how poverty may cause population growth."

The world then becomes dynamic, not static and the essential question becomes “what is the system?”, rather than who or which side is to blame.

The concept that a system can cause its own behavior through feedback is not an intuitive idea, though living with it in the real world sometimes seems it may be We ourselves are complex entities displaying emergent properties of life, consciousness (and I would assert free will), creating through our interactions with others the emergent existence of societies, cities and nations. We seem to live with the inherent complexity already underlying our lives and learning more about it doesn't need to impair our ability to live our lives. It is rapid, new and non-coherent complexity which makes us afraid of dealing with wicked problems, deferring them instead. More problems then arise because our decisions can be based on an assumed unchanging state of affairs and our own biases.

This is the end for now of the look at Donella Meadows’ “Thinking in Systems - A Primer”. Again, read the book. Up to this point, she has covered the basics and has provided enough material to return to the Jerusalem Vision Systems Practice project in the next post.



< part 2>
<part 1>