(image via Flickr user jonseidman88)
A new study analyzing 3.9 million English-speaking Facebook users has concluded that 71% of status updates are “self-censored” — that is, modified prior to posting. The research, highlighted by Slate for its incidental implications for privacy and data-collection practices at the social networking behemoth, was undertaken by Sauvik Das, a PhD student at Carnegie Mellon, and Adam Kramer, a data scientist at Facebook.
Though we wish the study’s paper, presented at the Seventh International AAAI Conference on Weblogs and Social Media, broke out the effect by age at a greater degree of granularity, the revealed demographic factors behind such self-editing are interesting. Das and Kramer summarize the key results:
Posts are censored more than comments (33% vs. 13%). Also, we found that decisions to self-censor content strongly affected [sic] by a user’s perception of audience: Users who target specific audiences self-censor more than users who do not. We also found that males censor more posts, but, surprisingly, also that males censor more than females when more of their friends are male. Additionally, we found that people with more boundaries to regulate censor more posts; older users censor fewer posts but more comments; and, people with more politically and age diverse friends censor fewer posts. [emphasis added]
Explaining that “a user’s ‘perceived audience’ lies at the heart of the issue,” Das and Kramer’s findings build a case for understanding expression on social networking sites as synthetically inhibited. Of course, the ‘performed,’ ‘edited,’ ‘self-censored’ (or what have you) element of self-expression is universal, but studies such as this one help attenuate the notion that identity-based online transmissions of text and (self-)image somehow give rise to a new kind of subjectivity.
Social network transmissions are audience-determined, just like a letter or telegraph or headshot or self-portrait, and subject to tailoring and revisions. In other words — shocker! — the lexicon of expression on new forms of media demonstrates the same key behavior that governed old forms of media, rather than catalyzing a novel paradigm for self-expression. And, as one might expect of researchers working for the entity providing the data being studied (Das held a summer position at Facebook), the paper’s painful word choice dramatizes the observed effect. Does anyone call editing a holiday card “censorship”?
