Systems and Organization 101
Simple, Complicated, and Complex
The universe is full of systems. They're so prevalent most of us know them when we see them. But a rough and ready definition couldn’t hurt. We can start simple and build outward:
A system is a collection of elements related in ways that work together.
For example, a solar system is a collection of planets around the sun. The Internet is a collection of computers connected by a set of protocols. A cell is a collection of smaller elements, such as the mitochondria, that work in cells that compose the organs of the human body. We could go on, but things don’t get interesting until we see the differences between systems.
Distinguishing among different types of systems can get confusing, if not dull. So, we’ll sacrifice subtlety and get to the point. Consider a three-part distinction: simple, complicated, and complex systems.
Simple.
There is little or no change, variability, or interconnection among the elements of a simple system. Cause and effect are clear.
(Builder, meet hammer and nail.)
Complicated.
There are multiple interconnected elements, but these are tightly coupled and work in a predictable, primarily linear fashion.
(Mechanic, meet the 747 aircraft.)
Complex.
As with complicated systems, complex systems have multiple interconnected elements, but the interconnections are difficult, if not impossible, to limn or predict in their totality. Non-linear relationships and self-organizing aspects indicate these systems.
(Mayor, meet Manhattan.)
Some examples are more intuitive than others.
The question of whether a system is complicated or complex is one we should consider carefully. For most people, the two words are more or less synonymous. For others, ‘complex’ just means a high degree of complication. But not here. One system can seem complex, while another will be complex.
747s and Grey Matter
A 747 is not complex in the sense we stipulated. That’s because, despite the myriad parts and electronics, the 747 only does a few things that boil down to this: Fly people around safely. Yes, it has many different interconnecting parts, valves, wires, and gears—probably more than we could ever memorize. It is certainly a complicated system. But each part relates to others in a predictable sequence of causes and effects.
We need something more to achieve complexity.
In a living brain, the relationships among elements are dynamic. A staggering array of neurons fire in patterns, such that behaviors at one level of description somehow give rise to different phenomena at another. In other words, excitations in gooey gray matter give rise to a person's conscious, waking life. The patterns of interaction among the neurons are mostly unpredictable. 85 billion neurons can fire in any number of ways, with varying degrees of strength, affecting any number of other neurons.
And they self-organize. That’s complexity.
Simple systems can appear complicated. Think of three long strings of Christmas tree lights. Pretend the lights are nodes, and the three strings’ power originates at one end and is connected to a live power strip. All three strings have several bulbs along the line and terminate in a bulb. If you were to drop all three strings into a tangle on the floor at once, they would seem complicated to a child who happens by. But you can untangle the strings and show that they are not complicated at all. In this case, we might consider the power source the super-ordinate node, and the bulbs all obeyed a simple command from the power source: "Light up."
Again, while three strings of tangled lights might look complicated, they are actually quite simple.
Hierarchies and Networks
What, if anything, does this have to do with human systems?
Let’s start with a simple system that comprises only two plusses as its elements. In this case, let’s say it represents an organization with only two people. We can represent these two people as nodes.
+ +
A line indicates interaction between the nodes, whereas its absence indicates no interaction. Arrows will tell us something about the nature of that interaction in our scheme. An arrow that points at another node means “potential command.” Without arrows, it means “potential collaboration or exchange.”
If one of the nodes has a super-ordinate role in the system, the other will have a subordinate role. That means one node issues commands while the other node carries them out (and perhaps reports relevant information back to the super-ordinate node). Such a relationship would look like this:
+—————>+
A basic hierarchy.
On the other hand, if both nodes are autonomous, then their interaction may be one of exchange and/or cooperation. We call this a basic network:
+—————+
The most significant difference between hierarchies and networks is the degree of autonomy any given node possesses to interact or communicate with any other node in the system. The more comprehensive the potential for node interaction, the more "networked" the system is.
The number of possible actions in a hierarchy is limited to those permitted by the superordinate nodes. In other words, you can only do what the person upstairs allows you to do. But in a networked organization, your actions are unlimited unless your actions fail to serve the mission. That means with networked organizations, such as holarchies (or holacracies), there is a far higher degree of autonomy among the nodes (or “roles”).
As a Role Lead, you have the authority to take any action or make any decision to enact your Role’s Purpose or Accountabilities, as long as you don’t break a rule defined in this Constitution.
Because networks can handle a level of complexity greater than what the busiest node (smartest person) can handle, some networks or holarchies can scale more effectively than hierarchies. When you combine self-organization with scale, you get an organization that has undergone a transition.
From command-and-control hierarchy to self-organizing network.
A Note on Coase
Remember, sometimes hierarchical organizations are a perfectly reasonable model in some circumstances. Ronald Coase, in his famous “The Nature of the Firm,” reminded us that the constitution of an organization depends on transaction costs, which might also get away with calling collaboration costs.
Introducing new social technologies (management models) and networking technologies (communications software) — can reduce transaction costs. That means where hierarchies once predominated, newer organizations can now shift power—and thus function and form.
Transaction costs help us understand that growing organizations must deal with increasing complexity. Systems that maximize node (personal) autonomy while reducing transaction/collaboration costs will do a better job reckoning with increased complexity.
Just as decentralization at the level of society can help people reckon with increased complexity, decentralization at the firm level can help partners deal with increased complexity.
