Because system thinking is reasoning about a question, circumstance, or problem explicitly as a system, our starting point for system thinking should be a discussion of systems. Few words in the modern English language are as widely applied or defined as the word “system.” The definition that we use in this text is given in Box 2.1.

The definition has two important parts:

1. A system is made up of entities that interact or are interrelated.

2. When the entities interact, there appears a function that is greater than, or other than, the functions of the individual entities.

At the core of all definitions of the word “system” is the first property listed here: the presence of entities and their relationships. Entities (also called parts, modules, routines, assemblies, etc.) are simply the chunks that make up the whole. The relationships can exist and be static (as in a connection) or dynamic and interactive (as in an exchange of goods).

Based on this part of the definition, what does not qualify as a system? If something is uniform in consistency throughout, it is not a system. For example, a brick (at a macroscopic level) is not a system, because it does not contain entities. However, a brick wall would qualify as a system, because it contains entities (many bricks and much mortar) and relationships (load exchange and geometry). Likewise, if a set of entities have no relationships (say, a person in Ukraine and a bag of rice in Asia), they do not constitute a system.

Notice how hard one must work to define things that are not systems! Someone might argue that at the right scale, a brick is a system: It is made of clay, which itself is a mixture of materials, and the materials have relationships such as sharing load and being in a geometric form (a parallelepiped). Likewise, a person in Ukraine could spend a euro to buy Asian rice, linking these entities into a trading system.

In fact, broadly construed, almost any set of entities can be interpreted as a system, and this is why the word is so commonly used. A closely related concept is the adjective “complex,” which (in its original and primary sense) means having many entities and relationships. In some languages, the noun “complex” is used to mean a system, as it sometimes is in technical English (as in “Launch Complex 39A” at the Kennedy Space Center).

Two ideas that are often confused are the concepts system and product. A product is something that is, or has the potential to be, exchanged. Thus some products are not systems (rice) and some systems are not products (the solar system), but many of the things we build are both products (exchanged) and systems (many interrelated entities), so the two words have become mixed in common usage.

Another closely related concept is architecture, the subject of this text. In its simplest form, architecture can be defined as “an abstract description of the entities of a system and the relationships between those entities.” [1] Clearly, the notion of a system (that exists and functions) and architecture (the description of the system) are intimately related.

System thinking emphasizes the second property listed in the definition of a system: A system is a set of entities and their relationships, whose functionality is greater than the sum of the individual entities.

This emphasized phrase describes what is called emergence, and it is the power and the magic of systems. Emergence refers to what appears, materializes, or surfaces when a system operates. Obtaining the desired emergence is why we build systems. Understanding emergence is the goal—and the art—of system thinking.

What emerges when a system comes together? Most obviously and crucially, function emerges. Function is what a system does: its actions, outcomes, or outputs. In a designed system, we design so that the anticipated desirable primary function emerges (cars transport people). This primary function is often linked to the benefit produced by the system (we buy cars because they ­transport people). Anticipated but undesirable outcomes may also emerge (cars burn hydrocarbons). Sometimes, as a system comes together, unanticipated function emerges (cars provide a sense of personal freedom). This is a desirable unanticipated outcome. An undesirable unanticipated function can also emerge (cars can kill people). As suggested by Table 2.1, emergent function can be anticipated or unanticipated, and it can be desirable or undesirable. It is also clear that more than the primary desirable function can emerge from a system (cars can also keep us warm or cool, and cars can entertain people).

The essential aspect of systems is that some new functions emerge. Consider the two elements shown in Figure 2.1: sand and a funnel-shaped glass tube. Sand is a natural material and has no anticipated function. A funnel concentrates or channels a flow. However, when they are put together, a new function emerges: keeping time. How could we have ever expected that sand + funnel would produce a time-keeping device? And how did two mechanical elements, sand and shaped glass, produce an informational system that keeps track of the abstraction called “time”?

In addition to function, performance emerges. Performance is how well a system operates or executes its function(s). It is an attribute of the function of the system. How quickly does the car transport people? How accurately does the hourglass keep time? These are issues of performance. Take as an example the human system shown in Figure 2.2, a soccer (or football) team. The function of all soccer teams is the same: the team members must work together to score more goals than the opponent. However, some soccer teams have better performance than others — they win more games. The team portrayed in Figure 2.2 was arguably the highest-performing team in the world in 2014 — the German national team that won the 2014 World Cup.

The first principle of system architecture deals with emergence (Box 2.2). Principles are long-enduring truths that are always, or nearly always, applicable. The principles we introduce will generally begin with quotations illustrating how great systems thinkers have expressed the principle. These quotations suggest the timelessness and universality of the principle. Each principle also includes a descriptive part and a prescriptive part (which guide our actions), as well as some further discussion.

There are other attributes of operation that emerge from a system, such as reliability, maintainability, operability, safety, and robustness. These are often called the “ilities.” In contrast with functional and performance emergence, which tend to create value immediately, the emergent value created by these “ilities” tends to emerge over the lifecycle of the system. How safely does a car transport people? How reliably does the hourglass keep time? How robustly does the German national soccer team win? How robustly or reliably will the software run? When a car breaks down at the side of the road, is it a mechanical “ility” problem or an embedded software “ility” problem?

The final class of emergence is so important that it merits a separate discussion: severe unanticipated and undesirable emergence. We usually call this an emergency (from the same word root as emergence!). Cars can lose traction and spin or roll. A soccer team could develop conflicts and lose its effectiveness on the day of an important match. Pictured in Figure 2.3 is a natural example of emergence: Hurricane Katrina as it bore down on New Orleans. The devastation from this system was enormous.

These emergent properties associated with function, performance, the “ilities,” and the absence of emergencies are closely related to the value that is created by a system. Value is benefit at cost. We build systems to deliver the benefit (the worth, importance, or utility as judged by a subjective observer).

In summary:

• A system is a set of entities and their relationships, whose functionality is greater than the sum of the individual entities.

• Almost anything can be considered a system, because almost everything contains ­entities and relationships.

• Emergence occurs when the functionality of the system is greater than the sum of the functionalities of the individual entities considered separately.

• Understanding emergence is the goal—and the art—of system thinking.

• Function, performance, and the “ilities” emerge as systems operate. These are closely linked to benefit and value, as is the absence of emergencies.