Just Dropped In (To See What System My System Was In)

There are all kinds of systems.  The American economy is a type of system, the natural ecology of an area is a type of system, the weather is a system, and the organization that we call a project or program is a type of system.  Given our penchant as a species for classification there are, among this list, some major groupings that can be discerned.

First among these is whether the system we are observing is a generally a natural one.  Social scientists in general, and economists and philosophers in particular, make this error all the time and the reasons for it become clear when we discuss them.  Ecological systems are natural systems absent human interaction.  The weather as I have described it above is a natural system, though weather is influenced by the heat effect of cities, among other identified phenomena.  The stock market is not a natural system.  The behavior of individuals in the socio-economic system is not a natural system.  There may be certain behaviors among humans that are common or predictive given particular stresses, rewards, and stimuli and these are part of a natural system.  But behavior in a particular social system is separate from an accurate description of human behavior in general.

Second, all of the systems I have named are either simple or complex.  What we mean by complexity is determined by observation in answering a “how” question.  In this case the amount of complexity is determined by the amount of information necessary to describe the system.

For example, inference that leads to algorithmic complexity is one way to identify complexity.  The scale that is being described also determines what we mean by complexity in any particular case.  Thus, when I listed “weather” above, what exactly did I mean?  Hurricanes (very important to understand in Florida), tornadoes (equally important for most of the United States), dust devils, rain patterns, or some other phenomenon of weather?

As a career sailor the controlling equations and models to predict wave height and frequency was linear wave theory.  Thus, a seemingly complex system actually could be explained by a model that simplified the individual processes involved.  These equations work quite well in most situations, especially when wave frequency and height is influenced by water depth and is wind driven.

But then there are exceptions and I experienced one personally.  In 1984 my shipmates and I encountered a giant wave in the middle of the Pacific Ocean.   At one time the U.S. Navy and other ocean going services would discount these reports as simply exaggerated sailor’s stories.  Certainly for my shipmates and I the 60% roll, the damaged equipment, and the twisted keel were evidence of more than exaggeration.  One of the larger ships in our task force suffered twisted decks and had to head back to port.  The widespread use of satellite observation has shown that these waves occur all the time.  How do we account for such phenomena?

The answer is non-linear equations and an understanding of the systems in play.  When we talk of “waves,” we are really being non-specific in determining if they are complex or simple systems.  Language oftentimes fails us.  For example, how do we account for shallow water wave patterns and the formation of barrier islands, inlets, estuaries, etc.?  Complex systems are involved and require an understanding of several independent actors to explain the pattern that results.  The same applies to the occasional appearance of a giant wave.  A tsunami, a different kind of giant wave, such as that was caused by the 2004 Indian Ocean and the 2011 Japanese Tohoku events, is largely predictable up until it reaches the shoreline.  Then special equations must be used to determine the height, force, and amount of land inundation.

Thus the number of and interdependence of the parts in influencing the system then come into play in determining complexity.  Certainly central limit theorem influences this determination.  This number needs to be sufficiently high to overcome central control that can be explained by a single, overriding and controlling event or actor.  This is why the equation for a tsunami is simple and the tsunami itself is not a complex system until it reaches the shoreline, but its affect on the shoreline and the predictive behavior of the wave on land then falls into the category of a complex system.  In our project, if there are few people, limited outside influences, and one single, controlling force or personality, such as a manager or corporate chain of command, then the test of complexity fails and we have a simple system that can be monitored since the complexity of the system is limited by the complexity of the controlling factor.

So ecological systems are complex, weather systems are complex, the socioeconomic system is complex, but only some projects are complex.

Finally, does the system evolve?  That is, does individual and collective behavior mutate and adapt based on some small event or collection of events?  For social systems this would include not only crowd behavior or individual instinctual responses, though they are also important, but also an indication of learning.

As part of a social system, complex adaptive systems include what we call projects (or any human organization) but not all projects are complex adaptive systems.  Some are clearly so, particularly in aerospace and defense, and space projects.  The conclusions in terms of the types of assessment systems and modeling that we apply become clearer given this insight.  For public policy, earned value management as a method of determining progress may be sufficient for simple projects.  More complex projects require other, more integrated, methods and models.