Which definition of complexity is more helpful for you, the one give in PMBoK ed6 (perception of an individual) or ed7 (characteristic of a system)?

Thomas,
My apologies for the long reply, but my head was churning all night over the statement: “Difficulty may feed into complexity” One could say I "geeked out" over the distinction, but the science of systems is what I do:

I would not use those words at all and I would say you have that precisely backwards. That statement is equivalent to effect leads to cause. Difficulty is not a component of complexity, but rather complexity is a component of difficulty.

The term complexity pre-dates formal PM by over 200 years and the word roots from the Latin essentially make it *the organization of intertwined things*. Etymology aside, what people often misuse the word to mean is essentially the perception of how much Work that causes, where Work may be defined in the classic physics terminology as:

Work = Force * Distance.

In that equation, complexity is really an element of Distance. Difficulty itself (the perception), can both be seen as the force required at individual intervals along the distance, or as the total area under the curve, i.e. the sum total of Work.

Putting my Systems Engineering hat on, if you were to decompose a project as a system of interconnected nodes representing performers of actions, edges representing communication channels, and interfaces where some processes combine information between nodes across the edges, the complexity is literally the organization of the interwoven items.

Complexity may be considered a component of Distance in the formula for Work a couple ways. Multiple things may occur simultaneously at individual nodes where complexity would represent horizontal layers of total work at each time interval (a sandpile chart).

I do agree with your statement that complexity is also often time dependent, as time is required to cross any distance. A process with many steps will then accumulate Work over time, where complexity could be seen as vertical slices of the total area in a graph of Work. Force (degree of effort) required at any time interval however, is a separate variable than the number of places and times where force must be applied.

How much force is required and the total energy output is where our perceptions enter regarding how complexity creates Work. This is where some people add meaning to the term to describe not complexity itself, but rather the perception of how much Work it creates.

In reality, you may have systems with a high degree of complexity, but very well ordered and easy to manage. A crystalline structure is a classic example from nature; complex yet elegant in its simplicity.

A far less complex system with more chaotic interfaces may be far more difficult to control. When the PMBOK provides their definition of communication channels, there is an edge between every 2 nodes and the number of channels grows exponentially. In PM, we can limit the communication channels (e.g. working through a chain of command) to reduce the overall difficulty through reducing complexity.

I also agree to an extent with what you say about increasing complexity may be beneficial. You can leverage complexity by picking the right nodes and edges to reduce the total work, even though the general perception of complexity is that it leads to greater difficulty. Well architected systems do this by design.

This is an example of where PMI and others takes terms and modify them for their own usage and brand differentiation, diverging from other bodies of knowledge where the terms are more precise.

My aim is not to be pedantic here either. Using different terms for the same thing creates a lot of confusion, particularly across language barriers, and is why mathematicians and engineers use very precise terms.

Best Regards,
Keith