Hidden Order: How Adaptation Builds Complexity (Helix Books) [Paperback]

Many books on fashionable current topics like complexity theory and complex adaptive systems are very lively in expressing the potentials of the field. This one isn't. Most books in these fields are either way over the heads of non-mathematicians, or just recount the story of the origin of the field.

This one is extremely modest and understated, but has the special merit of explaining the basic principles of complex adaptive systems in a way that any attentive reader can understand completely. It doesn't dwell on non-linearity, it just mentions it as one of the important principles that characterizes complex systems.

This stands out as not only an exceptionally clear description of the basic principles with simple understandable examples, but also a surprisingly dull read if you're used to popular accounts rather than texts. Going from the popular accounts of Chaos and Complexity Theory to this is a little like spending months reading Dr. Seuss' charmingly excessive rhymes, and then going back to "See Spot Run."

So it would be easy to miss what is so great about this book, that it actually makes the underlying principles of complex adaptive systems accessible to virtually anyone. Without the fanfare, without the hype, without the flashy graphics, Holland describes step by clear step how agents interacting with each other in certain ways that reflect 7 general principles end up organizing themselves into systems with their own properties.

Holland does not spend much time on thinking of applications for complex adaptive system models, he makes a brief, almost off-handed mention of the possibility of locating "lever points" where small strategic changes can make useful large changes to huge complex systems in real life, and the proceeds to describe his 7 principles. This is followed by a more detailed discussion of the agents themselves that compose adaptive systems, and finally some discussion of ECHO, a class of models that illustrates Holland's application of the principles.

There is no mystical reverence for emergence here, or any questionable speculations, this is a conceptual introduction you can can use to get a solid background, and then form your own opinions about the topic and its implications, and go on to read the more advanced books on complexity and systems thinking.

This is probably not a book that will inspire you about the importance of the potential implications of the field (which are great), unless you tend to get excited reading about rules and algorithms. It is a book that rather matter-of-factly describes what may well turn out to be the underlying structure of a myriad complex systems in nature, including ourselves and our social organizations.

"What enables cities to retain their coherence despite continual disruptions and a lack of central planning ?" he asks on the first page of the first chapter. That's one of the few hints of the amazing possibilities raised by a real-life model of complex adaptive systems in this book. But if you keep that possibility in mind, this simple straightforward though often uninspiring discussion of systems principles could be a very useful little hinge to open a very big door to a remarkable field.

However, if you are new to the phenomena of complex adaptive systems (CAS) or agent-based models (ABM), this might not be the best intro book for you. This is particularly true if you are wondering what a genetic algorithm is right now. I think you will get the most out of the book if you are already somewhat familiar with CAS and ABM as Holland does not dwell on illustrative examples. (Yes there are examples, but they are very short compared to other authors on this topic.) Because of this, I think this book will be rather dry and technical and non-intuitive for a real newbie. If you have no idea where to begin, try _Growing Artificial Societies_ by Joshua Epstein and Robert Axtell.

One final comment: for excellent in-depth look at the reiterated Prisoner's Dilemna model with genetic algorithms that Holland briefly discusses, read _The Complexity of Cooperation_ by Robert Axelrod. (Axelrod and Holland mention each other in their books.)