Thinking in Systems: A Primer

CREDIT: qimono via Pixabay (CC/2.0)

Every time the conversation turns to systems thinking, it’s prefaced by an apparently inevitable discussion of what, exactly, systems thinking is. At risk of starting a review of one author’s book with another’s definition, my favorite comes from Barry Richmond:

the art and science of making reliable inferences about behavior by developing an increasingly deep understanding of underlying structure.

Art and science lends an interesting dichotomy, granting dominion over most of the human experience in one hand while cheekily endorsing its discontents’ charge of “pseudoscience!” in the other. Inference, behavior, and understanding speak for themselves, and from increasingly deep the none-too-subtle allusion that a system taken to its logical extreme will run deeper than anyone cares to follow.

Definitions aside, Donella Meadows’ “Thinking in Systems: A Primer” is an introduction to it all.

First, systems: an “interconnected sets of elements that are coherently organized in a way that achieves something.” By Meadows’ definition, the mug of tea sitting next to me is a certain sort of system, a non-renewable reservoir of tea-flavored heat that’s being consumed slightly faster than it disperses into the room. It isn’t a very interesting system, set next to a nuclear power plant, say, or an advanced economy, but it’s somewhat easier to understand.

Whether my tea or Three Mile Island, all systems can be described by common abstractions. These are a present “stock” (in this case, of hot tea) and the “flows” that feed and drain it. The system’s behavior is controlled by feedback–either positive (“reinforcing”) or negative (“balancing”)–that may dampen or amplify changes to the system state.

Meadows uses these building blocks to present a menagerie of basic systems, what she calls the “System Zoo”. It’s simplistic but utterly sufficient to illustrate a serious cognitive challenge: while we’re well-equipped to reason about linear relationships–give X, get Y–we’re far less able to handle feedback and the mind-boggling behaviors it can breed. We want a simple explanation for a system as a whole. Finding it may be prohibitively expensive, if it exists at all.

The second half of the book takes a field trip out to the real world. Corporations, governments, and ecosystems are riddled with big, complicated systems. They’re self-organizing, resilient, and arranged into a hierarchy of subsystems that mostly take care of themselves. They mostly work, most of the time, and (except for the occasional academic curiosity) they’re mostly ignored while we go about our days.

The alarm bells go off when a system’s behavior diverges from our expectations. Nobody cares if my tea goes cold before I finish writing (OK, I care. Nobody else). But we expect capitalism to produce a lively stream of new ideas; when rent-seeking monopolists drive an economy to stagnation, we’re surprised by the system’s bad behavior.

We shouldn’t be. Even slight problems in a system’s design or operation can nudge it into a “System Trap” where it will no longer operate as intended. The Tragedy of the Commons is one familiar trap: when inadequate feedback about a system’s stock is coupled with an incentive for consumption, individual, rational actors will deplete it without a second thought. Policy resistance, the drift to low performance, and addiction are other examples of system failure.

Meadows closes with a hierarchy of tools for tuning system behavior. Focus on changing flows, not the stock; better yet to focus on the structure of the system itself, or the information governing its feedback relationships. This is how companies are organized, regulations set, resources managed, and societies built. We’ve practiced these exercises since long before “systems thinking” was a blip on the radar. Still, common language saves misunderstanding and frames future conversations about how systems should be built.

Perhaps it says something of the nature of systems–familiar, yet counterintuitive; neither neatly constrained nor easily analyzed–that Thinking in Systems is ultimately less of a prescriptive answer key than a collection of exercises. Readers with any background in complex systems may be frustrated by its shallow, math-free approach, but for everyone else, “Thinking in Systems: A Primer” is exactly that.

Hey, I'm RJ! For more learnings about software and management, find me @rjzaworski or sign up for my semi-regular newsletter.

Let’s keep in touch

Send me timely updates on software, product, and process.