This book explores how individuals can act, by following simple rules, in such a way that the total collection of such individuals forms a system providing resources, relationships, and behavior that would otherwise be inaccessible to the individuals acting alone. Johnson uses examples of ants, cities, software, and the Internet throughout his book to illustrate these points. This notion--emergence--is a characteristic of what he views as self-organizing, decentralized systems, and Johnson represents emergence as key for understanding how systems work and for instilling information-management power to the Internet.
The main strength of this book is Johnson's identification of emergence as a key quality that cities exhibit as they get large enough and healthy enough for people to interact for a variety of purposes.
Johnson makes the observation that Jane Jacobs' accomplishment in The Death and Life of Great American Cities was to tie the success of sidewalk life to emergent behavior. He observes that streetlife isn't important simply for the rich diversity it might represent, but for the crucial information exchange these encounters engender. People pass each other and the "constant succession of eyes" supports subtle social cues about expected behavior and conveys a sense of safety (pp. 93--94). Johnson further emphasizes how "left-leaning urbanists" can become mixed up in the value of sidewalk life for its "diversity" as an end in itself, while the strength of Jacobs' observations and Johnson's emphasis is that the small interactions of individuals adds up to large-scale intelligence, a "complex order" (p. 96) that acts to strengthen local commerce, local connections, and local problem-solving through the small-scale, fine-grained mix of people on sidewalks. Each person takes in many cues (about new stores, for example, or the condition of a neighborhood) and acts on these cues, and the aggregate behavior breathes life into a neighborhood or deadens its vitality.
Johnson also observes the emergent behavior of cities in terms of their function as "information storage and retrieval devices" (p. 108). In this metaphor, the city is a user interface itself, with clusters of goods and ideas and neighborhoods to help people make use of resources and relationships.
Johnson makes some apt observations about Internet-based information systems. He observes Alexa and Slashdot, and comments on the way information in each of these systems exhibit emergent behavior. Key to Internet-based information organization, he claims, is the ability for it to scale to the enormous amount of content online. Alexa uses a toolbar method to reveal Web page relationships based on automatic monitoring of user surfing. Slashdot uses a system of moderators in which moderators rate posts and moderators get rated themselves in a sort of rolling popularity contest. Johnson, writing for this book published in 2001, claims that "in a matter of years a Web page without a dynamic rating system attached will trigger the .... response... what's the point?" (p. 157).
I enjoyed this book, but I grew a bit impatient with its tone. A big burden author takes on is the task to make emergence seem like "the next cool thing" and market it to the reading audience by taking the stance that emergence is everywhere, could likely solve any problem, and is the wave of the future and what all the cool people are tapping into. I've no doubt that emergence is all of these things, but emergence is just one more key concept on a heap of key concepts that together mix in a kind of stew in which real life takes place. The self-organizing rules of emergence are logical and consistent, and human beings are not. And the rules are the central, static, hierarchical control that Johnson claims is not there. Indeed, the ant colony's queen does not busily dispatch and track her minion's work to accomplish milestones in ant hill development, but the ants do follow their DNA and the built-in instincts and pleasures. The artificial life software Johnson observes has programming rules created by human beings--programmers--that drive the "artificial life" onscreen, and so those software rules are the central control. Disparaging terms like "top-down" or "hierarchical" or "central control" or "static" and repeating buzzwords "dynamic" and "interactive" over and over seems somewhat irrelevant to the points Johnson makes about emergent systems, although this was the rhetoric of the dotcom boom that crested while Johnson wrote.
By claiming that there are no rules that act as a central, static, hierarchical, top-down organizing principle for a system, one obfuscates the true basis of how systems operate. Applied to cities, this obfuscation could mean that people observing suburban sprawl could claim it is a spontaneous expression of what people want, an emergent system (and therefore cool and modern and true and good). But a fixed set of rules make sprawl possible and alternatives impossible, and those rules are what drives the system (See Marshall and Duany).