January 15, 2010
Zuckerberg's Privacy Dispute: A Need for Comparative Social Network Analysis

If the Google v. China incident didn't steal all of your attention, you may have come across a short interview by Michael Arrington with Facebook founder Mark Zuckerberg (8 January), a few minutes of which deals with privacy on Facebook and across the social Web.


Watch the interview above, but the relevant content begins at 2:30 and ends at 4:00.

After the interview, many blogs went into a frenzy, proclaiming that Zuckerberg had declared privacy over and done with (eg., Facebook's Zuckerberg Says The Age of Privacy is Over, ReadWriteWeb). In two sentences, these were Zuckerberg's remarks:

People have really gotten comfortable not only sharing more information and different kinds, but more openly and with more people. That social norm is just something that has evolved over time.

As a general statement, we might declare 2000-2009 the decade of the social Web, in which a large sum of the general population entered the online space (versus the '90s and previous, which catered more toward computer scientists, specialized academics, and niche early adopters). With a new generation of users, then, the social Web defined the progress of the evolution of Internet culture: that is, how people interact with and are mediated by the technological infrastructure to produce or consume culture.

As I have mentioned before (Practical Geographies: Understanding How Cultural Practices Shape Social Media Usage), the Web Ecology Project has aimed to study the mediated space between users and platforms. We have achieved certain results by analyzing publicly available data over social networks (eg., Twitter, FriendFeed, etc.), while also scraping information from networks with privacy settings (therefore, we could only access that information available in our personal networks).

But the problem here is not a conflict over data for research. Nor is it a battle over keeping information away from companies. Rather it is a basic issue of providing the user with the ability to shape the platform according to his or her own preferences. It just so happens that the idea of social network produsage is commencing with the issue of privacy.

While the comfort level of general Internet users sharing information certainly has increased since 2000 (remember when most people were worrisome about using their credits cards on Amazon?), I am hesitant to agree with Zuckerberg's statement that users wish to share personal information more freely (in terms of volume and number of recipients). Certainly it's easy to see that the concept of spreading information across various networks has become a frequent practice. However, I would argue that users are currently more conscious about which information they share than at the beginning of the decade.

However, I will also argue that a user's understanding of and relationship to information depends heavily on how the user understands and relates to the platform which he or she uses. On one side of the spectrum, Facebook operates with user profiles which are interconnected with other profiles to create networks of "friends." Information can be "shared" when accounts intersect across friend networks. On the opposite end, a website like Craigslist.org thrives in user anonymity, where no user networks exist and where no personal information is shared between users (on the website, in theory, of course). In fact, it's positively "old-fashioned," as a Wired article puts it (Why Craigslist Is Such a Mess). "It relies on email and the telephone in an era of SMS and social networks. It sticks to traceless transactions in an industry that makes its living collecting data from every touch."

Of course, these two websites flourish based on the assumed necessity of sharing information. Contrastingly, OKCupid, the popular web-savvy dating site, allows users to set privacy preferences from account creation. As a dating website, users are probably more in-tuned to exactly what details they share about themselves. But OKCupid's matching algorithm -- which suggests other users to contact or avoid -- specifically utilizes shared information to make the matches (ie. the more questions you answer about yourself, the better the match).

Ultimately, the difficulty in debating about privacy is that each platform requires its own analysis. To understand the larger picture, therefore, more studies of cross-network analysis are sorely needed. The Web Ecology Project had attempted a study in the past, but we hit a wall: it was confusing to compare social networks without creating equivalency between different features on each network. danah boyd has written a few analyses, but they tend to share similar traits (eg., her study of status updates focuses on the two most similar networks, Facebook and Twitter: Some thoughts on Twitter vs. Facebook Status Updates).

In the end, we also have to remain conscious of the evolution of the social Web. When Facebook was only available to college students, users tended to share a lot of information and friend arbitrary people. But as Facebook has opened up to all users, these trends have significantly decreased, and it is common to even delete information before going to a job interview or censoring your profile before friending a family member. If Facebook's ultimate direction is toward open information practices on all ends, users will adapt to share less information, or at least similar amounts with smarter strategies in mind.