As part of the Virtual Book Tour, author Ethan Watters will be posting today at both Blogcritics and bitter-girl.com about topics that tie into his book Urban Tribes: A Generation Redefines Friendship, Family, and Commitment as a guest of Blogcritic Shannon Okey. See this Blogcritics post for details.
(See Chapter 2 of Urban Tribes: A Generation Redefines Friendship, Family, and Commitment for more background on these questions)
Shannon, bitter-girl.com: What is a "clustering coefficient," really? How do clustering coefficients differ when comparing in-person, organic tribe development with online "tribe-trackers" like Friendster? If a link forms in the woods and no one's there to see it, does it still exist? What's the difference between having a person in your extended social group who you don't much like (such as Jacob), someone you see in person and actually spend time with, and a person who keeps reappearing as your "friend" due to the sheer number of people you have in common online? You say:
My true relationship with Jacob could be understood only - and I mean only - through an understanding of the dozens of friendships that connected us.
What do you mean by "understanding"? Are you ranking the various friendships that connect you in some way? Are you looking at how each developed, or their current status as a means of determining how you got to this place where Jacob is a constant? How does that work?
Ethan Watters: The "clustering coefficient" is usually used to describe the interconnectivity of in-person groups. The most "clustered" group is one in which every one has a distinct relationship with everyone else. The most loosely clustered group would be one in which everyone knows one person but not anyone else. (Basically you divide the total number of relationships between people in the group by the maximum number of possible relationships that might exist between that many people.) The higher the clustering coefficient, the more a group feels like a group. I think it's a concept that could be well employed in assessing the meaning and importance of on-line communities. Do a lot of people know only a few people or do a lot of people know each other?
It's important to note that clustering coefficients only measures first-degree connections. The compelling part of Friendster and tribe.net is that we see the people outside those we know. The most interesting aspect, for me, is finding someone who's two or three degrees away from me but connected to me by a half dozen different pathways. Network theorists probably already have a word for this type of interconnectivity. I'd like to know what it is.
Article comments