Talk:Complexity
The discussion of complexity is totally wrong. There is no room for subjective complexity in systems engineering. Thre are plenty of sources of ciscussion of objective complexity. I suggest Sarah Sheard's Systems Engineering journal article for starters. There are plenty of objective characteristics of complexity. For example, the individual elements must be adaptable. Adaptability is an objective characteristic. I also suggest Dr. Scott E. Page's DVD course called Understanding Complexity. Dr. Page does not even discuss subjective complexity. He gives four characteristics of complex systems, all objective.
Hillary S adds:
I am very sympathetic to the above anonymous comment. I agree that a key issue for SE is objective complexity. However I don't think we can dismiss subjective complexity as irrelevant to SE; and the article needs to take a wider and more clearly structured view of the subject. There is a very lively debate with a number of different and apparently inconsistent worldviews between different parts of the community. SE does need to consider subjective complexity as well as objective, because if stakeholders have different and inconsistent understandings of the system and its purpose, due to unmanaged subjective complexity, the engineered system is unlikely to be successful. Unfortunately many of the proponents of subjective complexity do not believe that the objective kind exists. It seems to me that this section needs to explicitly set out and recognise a number of different views of complexity, non-judgementally, and show how each is currently believed to relate to SE practice. (My papers [Sillitto] from IS2009 and 2010 try to give a clear view of subjective vs objective complexity, which is a different dimension of distinction from Senge's behavioural vs detail complexity types.)
The Aslaksen structural complexity definition looks useful and the equation presented in the text should be formatted more clearly.
Scott Jackson adds:
The previous comments were not meant to be anonymous. Some comments by the organisar led me to believe that contributors would be automatically identified.
There is no doubt that complex systems are hard to understand. My point is that subjective complexity cannot be treated in systems engineering. Subjectivity is an interesting side note. From memory Scott Page's four characteristics are as follows: (1) All the elemens of a complex system are independent. Take a flock of geese. Every goose can make its own decisions. This is not a subjective fact. (2) Elements of a complex system are interconnected. Every goose knows what the other geese are doing. (3) Elements of a complex system are diverse. Their decisions can all be different. (4) Elements of a ocmplex system are adaptable. Each goose will do what is best for itself. All the above are objective characteristics.
One added point. Objectively defined complex systems can be modelled. The common form of modelling is calle agent based modelling. Subjectively defined systems cannot be modelled.
I think the emergence discussion is concise, useful, and sufficient for purpose. [Hillary comment]
Scott comment: Subjectively defined systems do not have emergence. Emergence is aother objective characteristic.