Sub-rosa assumptions of community tools

I posted a link to Clay Shirky's Many-to-many post on moderation on the right, but thought I would pull this bit out because it's so good:

I love this post, because it articulates what I think of as the sub-rosa assumptions around earlier forms of community tools:

- Systems should only use technological, not social, tools
- A user is responsible only for his or her own behavior
- Any policy to be enforced must be expressible algorithmically — no judgment calls
- Users must have access to pseudonymous communications

The central thesis of the post — that sponsorship can’t work, for these reasons — is suspect at the very least, as sponsorship systems work well elsewhere, and humans use both social influence and judgment calls to affect one another’s behavior for some time now.