Inventing Elephants

Most natural systems recover well when buffeted by winds of change, even when they come fast and from many directions. But then when we try to control the results in the short term or define a too narrow range of performance, we can defeat the very mechanisms that allow them to behave so well.

Resilience brings the system back to center.

Oscillations, or conditions swinging back and forth around a central point, are often normal, both in the short-term and the long term. A healthy system has a collection of balancing loops that operate in different ways across these different lengths of time that bring the behavior back to that central point.

If we look at the system in too short a time frame or if the swings in behavior are too extreme for our expectations, then it sure doesn't feel like it is behaving well! Then we might try to keep the recoveries from occurring, but without resilience the system would really take off in a direction we had not protected.

Self-organization allows a system to become more complex on its own.

Snowflakes grow and communities come together by following a few simple rules and relationships. But these simple rules make possible a larger, more complex system that can meet a challenge or just a slightly changed situation.

However, this can be disconcerting to us, because there's no way to tell from the initial state and the rules (which we don't always know) exactly what the end result is going to be. This is how we come up with snowflakes that all look different. If we try to say that only certain snowflakes are an acceptable shape then we keep the system from being able to reach its full potential.

A hierarchy of subsystems keeps effort from being repeated.

Most systems are layers of systems within systems, like cells in organs in our bodies. Because different functions can be assigned to different systems, like the lungs or the liver, then the overall system functions more effectively. In addition, two subsystems in different areas often don't affect each other directly.

Hierarchy is also why the more usual practice of studying things by breaking them down into their parts works. We just can't forget that the parts interact when they are in together. 

This post is the fifth in a series that discusses the concepts in Thinking in Systems by Donella Meadows. It summarized chapter 3 which is titled rather more optimistically "Why Systems Work So Well." I found the comments about short-sightedness that she embedded into her stories to be as important as the three characteristics themselves and tried to include a sense of them here. Also read my other posts:

1. Book Review: Thinking in Systems
2. What is a System?
3. Feedback Entangles How Fast with How Much
4. Delays and Disasters at the Zoo
5. Effective Systems Beyond Our Control
6. Why Systems Surprise Us
7. at least four more to come....

"A Brief Visit to the Systems Zoo" is the appealing title of chapter 2 in Thinking in Systems, and is intended to show that Donella Meadows will be isolating and looking at small interactions by themselves instead of in their larger ecosystems.

What follows is merely a summary, with one key point taken from each detailed example. The book goes through them step by step including visualizations of the system itself and its performance over time.

There will always be delays in response.

A thermostat is a stock with two competing balancing loops, the flow of heat from the furnace and the flow of heat to the outside. You can only make changes that affect future behavior, never current, and when you need to remember to make one change go fast enough to correct for the actions of the other.

One process can have a stronger effect than the other.

Population growth and decline is a stock with one reinforcing loop - fertility - and one balancing loop - mortality. Depending on the circumstances, one loop will be dominant and have a greater effect. This is also often true in more complex systems. Just because interactions are present does not mean that each cycle is equally important.

Lengthening or shortening delays can have a profound effect.

The changes in business inventory is a system with multiple delays.

Delay in perception - how long it takes to decide that an increase in sales means an order needs to be made.

Delay in response - how much of the shortfall you try to order at one time.

Delay in delivery - how long it takes to receive the product.

This creates an oscillation when the the inventory flip-flops between being overstocked and understocked. The human reaction to this annoyance is usually to react more quickly.

Acting faster actually makes the situation worse!

But lengthening the time it takes to decide to order and increasing how much you order at once allows you to have a more stable inventory level.

I'm still trying to wrap my head around this one...

It's because she is struggling to operate in a system in which she doesn't have, and can't have, timely information and in which physical delays prevent her actions from having an immediate affect on inventory. She doesn't know what her customers will do next. When they do something, she's not sure they'll keep on doing it. When she issues an order, she doesn't see an immediate response. This situation of information insufficiency and physical delays is very common. (emphasis mine)

The higher and faster you grow, the farther and faster you fall.

At least in a system like the oil economy, which has a renewable stock constrained by a non-renewable stock. This is also a reinforcing loop and two balancing loop, but the difference is that there's a physical limit to "how much" versus "how fast".

There is only so much oil in a field, for example, the non-renewable stock, but there is potentially a limitless and renewing supply of capital money to extract that oil. At least until the oil runs out. And, as Meadows notes, the system still works this way if you factor in technological improvements that allow you to extract more oil.

Effectiveness of renewable resources depends on the system too.

In contrast, the fishing economy is a renewable stock constrained by a renewable stock. Again, capital money is one stock, and the size of the resource, in this case the amount of fish is the other. There is also still a reinforcing loop and two balancing loops.

The system will almost always overshoot, followed by one of three results:

• a sustainable equilibrium - we can still eat fish, just maybe not as many at once
• oscillations around that equilibrium - prices and supply bounce around and we might go to
• a collapse of the resource - no more fish

Just because the resource is renewable, doesn't mean we're safe. We have to avoid taking the level to the point below where it is capable of regenerating. And we have to have a balancing loop that is strong enough to slow capital growth as the resource decreases.

This post is the fourth in a series that discusses the concepts in Thinking in Systems by Donella Meadows. Read the book to see her diagrams, graphs, and capture more of the subtleties of the concepts, especially when it comes to this chapter. I think I still need to reread it a few more times to get everything I can from it. Also reach my other posts:

1. Book Review: Thinking in Systems
2. What is a System?
3. Feedback Entangles How Fast with How Much
4. Delays and Disasters at the Zoo
5. Effective Systems Beyond Our Control
6. Why Systems Surprise Us
7. at least four more to come....