As it turns out, I haven't been the first to use the word balkanization in describing social networks' tendency towards segmentation. David Frank VanDevelder has discussed the balkanization of knowledge recently, and Pacey Foster has a really neat SNA take on the balkanization of influences in reggae music sampling. Prophetically, however, MIT's Marshall Van Alstyne and Erik Brynjolfsson suggested way back in 1996 that "improved access, search and screening might either balkanize or integrate interaction" within the rapidly-emerging global community.
Their essay, Electronic Communities: Global Village or Cyberbalkans?, points out that improved communications technology doesn't necessarily guarantee a sudddenly-unified global village. By means of numerous algebraic operators I've had to look up, Van Alstyne and Brynjolsson suggest that the very impact of social software is socially-defined. So perhaps a unified network might emerge if users prefer infinite heterogeneity, but improved technology could increase balkanization where users prefer more focused interactions than they currently have. "Thus local heterogeneity can give way to virtual homogeneity as communities coalesce across geographic boundaries," they argue.
Now, I've argued in the past that a unified profile space should emerge because it would be globally optimal. Van Alstyne and Brynjolsson note that human constraints of Bounded Rationality limit individual tendencies to act in globally optimal ways. Indeed, as I've also noted, it's sometimes optimal to exclude irrelevant information.
Out there in the world of social software, though, platform designers still seem to expect global egalitarianism despite theory and trends to the contrary. Paul told Doug and I just a few days ago that Meetro won't be targeting any particular user groups at the outset. TheFaceBook might be argued to fit this description, although I understand it will be available to literally every four-year college student by the beginning of the fall semester in the next couple weeks. MySpace and LiveJournal still don't seem to have met users they haven't liked. So let's look briefly at a micro-network with a large but clearly-defined audience.
Rediff Matchmaker is an Indian matchmaking site that bills itself as the place "where progressive Indians find their match." Users may search to find either a bride or a groom along such dimensions as age, religion, language, photo and educational background. It's not clear just how many users have submitted profiles to the site, but the homepage highlights four recent Rediff marriages. By tapping into the long-standing Indian tradition of arranged marriages, Rediff might well attract many descendants of the world's second most populous country. But in fact, this site's basic value proposition arises in part from the massive swath of humanity who don't belong.
I'm sure Van Alstyne and Brynjolfsson's research will yield more interesting insights as I study it further in future weeks. For the moment, however, it's clear that there lies crucial social information within the interactions of homogeneous groups. I still believe that the best long-term equilibrium would include a unified profile space upon which overlapping network sites could draw. In the nearer term, I'm going to keep a close eye on networks with well-defined target audiences.
Always happy to contribute to the conversation. For some reason, my paper has been moved from the above link; please find it at the following:
http://www.cyberianpost.com/vandi/resume/thesis.html
Thanks again.
DFV
Posted by: David F VanDevelder | May 09, 2007 at 12:53 PM