Turtles, Termites, and Traffic Jams: Explorations in Massively Parallel Microworlds
It never really occurred to me that bird flocks don't have an "alpha" bird leader. In fact, most of the behaviors observed in the animal kingdom are decentralized; that is they are not controlled by a "central controller." This fundamental observation extends well into human systems as well. The economy is a complex interleaved set of human behaviors. When things go wrong, we all too often try to pin the cause on one or two individuals, or groups of individuals. Likewise, many politicians like to take credit for a well performing economy under their watch, but the vast majority of evidence suggests that the overall health of an economy is based on the system itself, and not any one or more individuals, no matter how authoritative their position may be. That is simply the nature of human systems.
Turtles, Termites and Traffic Jams introduces us ways to understand the tendency for systems to become decentralized, and demonstrates tools to observe phenomena through simulation and models. A classic example is the study of a traffic jam. Give simple rules like, "If there is a car close ahead of you, slow down" and "If there aren't any cars close ahead of you, speed up" a traffic system can be simulated. Cars independently slow down and speed up given their own control. However, in an environment where there are constraints, say a police officer with a radar gun, then chain reactions start to emerge. Cars going too fast must react to this condition, thus causing cars behind to slow down when there otherwise would be no apparent reason to do so. This emergent behavior can be described as complex.
Though the author organizes this book into five sections, the book is essentially divided into two parts; the first (Foundations) delivers the underlying principles to decentralized, self-organizing and emergent behavior. The second part covers examples of various phenomena demonstrated via simulation. The result is a thoroughly interesting, though provoking book that leads the reader to want to try the simulations, if not on a computer at least in thought. The subjects of the simulations, often times natural systems such as bugs, frogs, turtles and of course termites, are less important rather than the underlying phenomena that's being explained. Artificial Ants, for instance, demonstrates problem solving skills among a society of near clone-like creatures whose singular goal is success for the colony, rather than individual success. They are able to find food sources, and communicate important information to exploiting those food sources to the fullest. For example, if they find an apple core on the ground, does the whole colony attack it? How many ants should be out looking for more food sources? The answer lies within the mechanics of communication. Using a pheromone, an ant "smells" its way to collect sufficient information to steer it in the direction it would best serve the colony. When a food source is becoming depleted, less pheromone will be apparent causing a greater proportion of ants to go back out and scout for more food sources.
The problems presented, while not in a business context, do have important implications. How can businesses succeed in difficult competitive environments with little information? The demonstrations of systems observed in the natural world give not only inspiration, but important clues to how businesses may survive in the competitive market world. I highly recommend to anyone who is interested in agent-based modeling and the principles it is based upon.
Resnick, M. (1994). Turtles, termites, and traffic jams: Explorations in massively parallel microworlds. Cambridge, Mass: MIT Press.
Found available at: