I just finished reading this book with a really long title: Connected: The Surprising Power of Social Networks and How They Shape Our Lives. Whew. Being a connoisseur of social network research, this book was very relevant to my interests. Previously I had read Six Degrees by that Duncan Watts guy.
In Connected, Christakis and Fowler argue that social networks are a natural part of our lives. We’ve evolved to be good at navigating them and using them for our needs. While it’s simple to prove our actions have an effect on people who we are directly connected to, we apparently also have an effect on the people who are connected to the people we are connected to. This effect fans out from us to our third degree contacts (friends of friends of friends). Our actions might affect others’ obesity, happiness, loneliness and promiscuousness. Conversely, people we don’t even know can affect whether we’re happy or fat (not mutually exclusive).
I actually saw James Fowler talk about some research that made it into the book at a complex systems seminar at the University of Michigan back when I was a student. He focused on how happiness could spread and talked less about the fatness issue. He apparently caught a lot of flack for writing about how the obesity “epidemic” really is an epidemic. That is, the more fat friends you have, the more likely you are to become fat. You could imagine how this could make overweight people feel: like some vector for a contagious disease known as chubbiness. While the oversimplification seems funny, when you boil it down to human behaviors (as the book describes), it makes sense. People mimic each other. When everyone else around you is overweight, it’s easy to assume that behaviors that lead to weight gain is the norm. It works with lots of other things too. And that’s basically what the book amounts to. Many examples of people in networks spreading ideas, diseases, love, hate, etc. The book also touches on how technology has amplified our natural ability to network into hyper-networks.
One criticism of the book I have is that it feels a bit too dry. The book is describing all kinds of interesting phenomena, but it does so in a really scientific way. Don’t get me wrong, I like science, but I also like a story in my non-fiction hardback book. Six Degrees is a good example of a great non-fiction book that tells a wonderful story. The narrative fits together well and the content is interesting. Duncan Watts is good at both science and storytelling. You could compare this against Malcolm Gladwell, who is good at telling a story but stinks at actual science. He kind of makes stuff up and jumps to conclusions that aren’t really backed by anything factual. Christakis and Fowler (which one did more writing?) are great at the science but the story lacks a clear narrative. Extending the idea for Pasteur’s Quadrant, Gladwell is in the “only story” quadrant, Christakis/Fowler seem to be closer to the “only science” quadrant, and Watt sits pretty well in “story + science” quadrant. Freakanomics authors Dubner and Levitt seem to fit into “story + science” though sometimes I feel they like to jump to conclusions like Gladwell does. If I were a researcher-cum-book-writer, I think I’d like to sit in that Watts quadrant myself.
Overall, I think the book is compelling and interesting. The ideas from the book have already gotten my head juices flowing and thinking about future network analysis directions. Connected could definitely be improved by including some underlying narrative or cohesive theme, but it’s still worth reading as it is.
Next on my plate is “Linked” by the guy who was named after that preferential attachment network model thing. Or was it the other way around? I just realized that the set of names for these network analysis books is growing smaller day by day! We’ve already seen books called “Linked,” “Connected,” “Nexus,” and “Six Degrees.” I wonder what the next one will be called? “Affixed?” “Conjoined?” “Hooked Up?” Maybe I should claim the name of my book before someone else does.