*I'm reading Donella Meadows' "Thinking in Systems" for some research I'm working on, and summarizing it for a few other people I work with. Since the ideas in the book fascinate me, and I need to get back to blogging (mostly this), I'll write several rambling, nuanced, directionless summaries of what I read and post them here.
What makes a system?
-elements
-interconnections between elements
-a shared purpose
Many entities fit the description of system: the human body, baseball teams, multinational corporations, trees. All are made of elements that work together to accomplish something. Non-systems are inanimate, and do not do anything of their own accord.
-Elements are not always tangible or easily observable. For example, school pride and academic prowess strongly shape a university system.
-Interconnections are more descriptive of the functions of the system than are the elements. A tree, as a system, is a fitting example: Although the tree is composed of roots, a trunk, branches and leaves, its growth and use of nutrients is the result of a complex set of exchanges and signals from roots to leaves that act to distribute and make use of the nutrients. Although defining the parts is essential, understanding the relationships between them is more insightful as to a system's productivity.
-In order to adapt to shifts in environment, interconnections adapt and sometimes reshape the elements. For example, in a tree system, as temperatures and seasons change, trees put forth chemical message that cause nutrients to migrate out of the leaves and into the trunk.
-interconnections can be actual physical flows, but are often characterized by the flow of information
-Purpose is deduced from behavior, not stated goals or rhetoric. Thus, it is often more insightful to observe what a system does and describe it, rather than to talk about what it says it does. For example, if we see a frog look right and catch a fly, look left and catch a fly, and turn around to catch a fly, the purpose of the frog is not about turning right or left, but, rather, catching flies.
Changes to the System's Behavior
-System purposes are not necessarily values held by the actors within a system. There are also systems often nested within systems. Keeping sub-purposes in harmony with overall system purposes is an essential function of successful systems.
-A change in the behavior of a system is not brought about by a change in one or more elements alone. The example is given of football team, where players are constantly traded for one another, coaches come and go, and new equipment replaces the old. But the game goes on, business as usual, until some relationship, whether between player and coach or some other set of elements, changes.
-Changes in purpose also dramatically affect outcomes. In the case of the football team, what would happen if one or more of the elements decided that its purpose was losing not winning.
-The example is also given of a change political leadership. A new President may be elected to replace the incumbent, but whether or not the presidency changes his country and his people depends on whether or not the nature of the interrelations of the Presidential office change, and they often do.
Stock and Flows
Stock refers to elements of a system than can be seen, felt, or measured at a given time. Population, money in a bank account, wood in a tree--these are all examples of stock. These are typically characterized by their ability to flow--interconnections--within the system. Within a health context, we might call the number of beds in a hospital stock, as well as the number of patients occupying the beds. Human or financial resources are also stock.
Understanding the dynamic of the flow of stocks is vital to understanding the behavior of a system. For example, if we want to fill up a bathtub, and our variables are input from the faucet and output through the drain, we take into account how the water coming in and the water going on must related in order to reach the intended goal. It is important to note that turning up the faucet is not the only way to cause the water to rise; partially or completely stopping the drain can accomplish the same thing. The point is that progress is often characterized an interrelation of flows.
-A stock takes time to change, because a flow takes time to flow.
Changes in system behavior are never immediately affected as a result of changes in stock.
Feedback
"Systems thinkers see the world as a collection of "feedback responses."
-although the respective elements of a system are affected by flows and changes in stock, there is usually a delay between flows and systemic effects. Stocks change when flows are manipulated, but they change slowly, so that they act as buffers against radical systemic change.
"If you see a behavior that persists over time, there is likely a mechanism creating that consisten behavior. This pattern is called a feedback loop. Rules and laws tend to create feedback loops, as well as excesses and demands. An example of a feedback loop might be your personal checking account. As you spend money during the month, the stock (your money) decreases in a way that when you see your bill, you determine that must work a certain number of hours in order to replenish the account. The amount worked refills the account to allow for a certain amount of freedom to spend again. The loop here is regulated by the cost of living, your wages, the enjoyment you experience at work and in spending your money, and a near infinity of other factors. So long as they remain constant, it is likely that you'll continue going to work, getting a check and spending it. The level of stocks (money) do not inherently affect flows, insofar as they do not impede or enhance a desired end.
There is an interesting line: "Remember that all diagrams are simplifications of the real world." This reminds me that when we're talking about designing a maximally effective health system--that is, one that meets the needs of patients--the degree of complexity to which attention can afford to be paid ought to be considered. The infinite complexity of every given relationship, causes that we must simplify them into terms we can relate to. Also, the experiential, unknowable aspect of things--the way humans perceived them, as shaped by sociocultural constructs--has a great deal to do with the system's performance while it is much less measurable or predictable.
A stabilizing or balancing loop opposes changes in flow in order to keep stocks within desirable limits. We see a "homing behavior" in balancing loops, which seeks to bring the system back to its natural, most stable state. The example is given of the coffee drinker, whose desired energy level shapes the amount of coffee she drinks and the frequency at which she drinks it. When she is energized, she drinks less coffee to allow her system to function at a desirable level. When she is tired, she drinks more coffee to oppose her body's decline into lethargy.
Feedback--which is manifest in interconnections and the exchange of information or services--can often be misdirected or arrive too late, causing a weak, resource-constrained, or simply ineffective response.
A reinforcing feedback loop enhances whatever direction of change is imposed on it. An example is giving of two children beginning to fight. One pushes the other, the other pauses then pushes back, the other pushes back more quickly and soon they are wrestling. Another example would be that the more time one spends practicing the piano, the more pleasure one gets out of hearing themselves play. The more they enjoy playing, the more they practice, and so on. This is related to the concept of exponential growth.
Instead of asking who or what is to blame, the better question is "What's the system?" The concept of feedback emphasizes that a system can cause its own behavior.
