Friday, August 05, 2005

Books: Six-degrees: the science of the connected age

A good book came my way recently, providing a layman's explanation of the emerging field of network theory as it is applied to sociology. Six Degrees: the science of a connected age, written by Duncan J. Watts is not only a good introduction to a revolutionary style of analysis that is being applied to every field of science, but it also can help social activists to think about where they want our society to go and how to get there.

The basic idea is that social networks influence human behavior, which in turn influences the structure of these networks. The analysis of social networks begins by simplifying the network into "nodes" and "edges." Nodes represent individuals in the network (a person, business, city, etc.), while edges represent a connection between two nodes. An assortment of simple rules can be used to create the network, and then questions can be asked about the characteristics of the resulting network; for example, "what is the shortest path between two randomly chosen nodes?"

One interesting section addresses the "preferential growth model" of networks. In constructing this model, new connections are preferentially added to the nodes that already have many connections. In other words, "the rich get richer." This means that small differences in the early condition of the various nodes--whether due to inherent traits of the node or pure chance--can result in huge differences in outcome. This produces a "scale-free network", which shares some important similarities with real-world phenomenon including the size distribution of cities, the size distribution of business firms, and the distribution of wealth within any society. (for more info, see the Gini coefficient and Zipf's law)

There is also an interesting section addressing the structure of corporations. Many of us assume that a vertically integrated hierarchy is inherently more efficient than other structures simply because we have never seen any other structure used. Network theory suggests that there are several advantages to decentralized corporate structures, and Watts uses the Toyota-Aisin crisis as a case study of how decentralized structures can be more robust than centralized structures. He even goes on to show how this robust structure can arise naturally from the day-to-day activities of workers who are allowed to coordinate their activities without involving management.

This observation, that social structures are naturally modular and self-organizing, strikes at the heart of political prejudice that I really hate--that as a society grows in size and complexity, the government must become involved in more aspects of the society. I have seen this prejudice asserted without support more than once, including in a political science textbook. This strongly conflicts with my belief that no one is more fit to look out for a person's interests and evaluate his behavior than that person and his close associates (i.e., friends and family). I believe that network theory can provide us with the tools and intuitions to evaluate how our social relations should be structured. It can also help us to live in a society where there is no all-seeing, benevolent master watching over us and directing us.

Finally, this paradigm can provide social activists with a key for how to influence social decision making.
"Whether compensating for lack of information, succumbing to peer pressure, harnessing the benefits of a shared technology, or attempting to coordinate our common interests, we humans continuously, naturally, inevitably, and often unconsciously pay attention to each other when making all manner of decisions, from the trivial to the life changing."
Watts introduces a bit of psychology regarding how individuals make decisions, and then examines how the structure of social networks interacts with human psychology to produce our behaviors. These models emphasize the importance of having someone who will state unusual opinions, even in the face of popular opposition, and how this is the beginning of saving humanity from "the madness of crowds". It also provides some ideas for how social networks can be leveraged to increase the influence of one's own ideas.

Overall, this is a very informative and well-written book. I like that Watts takes the time to make a distinction that is often overlooked in our society: "there is a difference between holding someone accountable for their actions and believing the explanation for those actions is entirely self-contained." There is also some great writing in this book, such as "Listening to Chuck (Sabel) think is like drinking wine from a fire-hose."

By the way, the ideas of this book are closely related to the ideas that prompted me to begin this blog.