Annalee Newitz held a great session talking about how users can censor other people’s content at Web 2.0. Here are my notes from that session.
Social media censorship
- Bottom up, not top down: not imposed by an authority figure, like traditional censorship. Users tell each other what they can / can’t say and reporting it to the owner by flagging.
- Collaborative: groups of people work together to censor content that they don’t agree with or like.
- Punitive (cruelty of crowds): censoring content because a few users don’t like it even though it may be within terms of service and OK for the site otherwise.
- Not within terms of service: This really isn’t censorship. This is an appropriate reason to flag content.
Why do I Care?
- Censorship makes user-generated content less valuable
- Creates divisiveness w/in community
- Drives community away
- It is unjust
Let’s collaborate to destroy free expression!
Blogger: Flag blog. Annalee asked someone trying to get a job for her to start a blog, but it got flagged by another user. Blog got shut down, she was unable to modify it, and it was a long process to get the blog unblocked. It could have cost the girl her job.
Flickr: Flag photo. Violet Blue had her photos flagged as unsafe and her account reclassified as restricted even though most of the photos were fine. You can have a person review it, but there is no phone number and no time estimate for when it will be reviewed. The process to remedy user censorship is just not very helpful.
YouTube: much more granular lists of reasons to have something flagged, which helps them respond and forces the user to be specific about why something is inappropriate. They will act w/in 24 hours in certain cases (filmed murder, etc.)
Digg: There are a lot of debates about what it means to “bury” a story. It isn’t transparent. Creates controversy, since some groups use it to prevent stories that they don’t like or don’t agree with from going to the front page.
Wikipedia: They have very elaborate rules for content, which makes it harder to censor (unverifiable claims, references & sources not properly cited, etc.) You are less likely to see censorship, since there are so many rules around it.
Solutions:
- Clear content guidelines
- Clear and fast methods of redress when censorship has happened
- Easy ways for readers to use filters that prevent them from stumbling across content that upsets them.
Crowds can be wise, but they can also be destructive.

Blog RSS Feed (recommended)
Subscribe to Fast Wonder by Email






OpenID at ClaimID






“Wikipedia: They have very elaborate rules for content, which makes it harder to censor (unverifiable claims, references & sources not properly cited, etc.) You are less likely to see censorship, since there are so many rules around it.”
I’m not sure which Wikipedia Annalee was using. I quit contributing content to the Wikipedia I had tried after I saw one too many entries, including some of my own, redacted or deleted by rules-lawyer vandals in service of some bizarre ideal of “public good”, with the explicit cover of the management.
I think that “bottom-up censorship” is a poor description of this phenomenon, as the top-down management practices are the root cause of the listed social problems. Notice that the listed examples are all actually examples of bottom-up interacting with an ill-conceived content management plan. For a positive example, I’d suggest looking instead at a site like tvtropes.org which operates in the true “wiki way”, with almost no explicit content management at all.
It turns out that giving true power to the people to control the content of their informational spaces tends to work better than authoritarian rules. Spammers and trolls flourish not because the polis has too much control, but because it has too little. When you actually give the community the power to police itself, as true wikis and some muds do, folks tend to judge communications according to their intention rather than according to their content. You want that first thing.
Thanks much, Dawn, for blogging this stuff!