Which might wash, except that one prominent Digg member who was banned, DiggBoss, wrote a script that (a) did not submit stories and (b) enhanced Digg.com's functionality. (In short: Digg lets you e-mail friends to share stories you think are cool. Nothing in the existing software prevents someone from sending the same "shout" to recipients a dozen times. DiggBoss's script removed the duplication if the friend already dugg the story.) The result? DiggBoss was told he was banned for life. Gosh. That's a long time. You think that's maybe a little excessive?

Now mind you, it's Digg's party and they can make all the rules they want. Digg has plenty of challenges (many of which they've created for themselves, in my humble opinion, though that's another issue). But from a community management point of view, this situation is a major failure. The banning happened without conversation, without warning, without any opportunity for the community member to respond... even when (or maybe because) it's someone who was an active and committed participant.

It's fine to moderate the content of a community; as CIO.com's Advice & Opinion Blog Mom, I do it every day. It's my duty to ensure the real participants are protected from the aforementioned cranks and aggressive self-marketers. But it is a real human who makes the decisions and who listens to would-be contributors. Sometimes they are misguided about appropriate behavior, and a few words of explanation sets them aright. I firmly believe that ignorance is curable. Unless it's unrelenting spam, everyone gets a warning.

This is not always an easy call. I've been on both sides of the keyboard in these sorts of situations. For example, the right thing to do in user disputes is usually for the moderator to keep out of it and let people work things out for themselves, including letting users declare one another devil-spawn. Community members don't always get along; sometimes (such as political forums) that's part of the appeal. Every message-reading client since Tapcis for CompuServe has had an option for "ignore this user," and that's fine.

However, social networks, and any community that enables voting, adds a new and occasionally uncomfortable twist. Now, instead of a one-on-one grudge match, your online enemy can vote-down a contribution you make to reddit, for instance, or mark an Amazon review as "not helpful" (when they actually mean "I disagree"—common with political or religious books). Within bounds, that's okay.

But then there's the strange "stalkers," which we active Amazon reviewers call "neginators." As top-70 Amazon reviewer Duffbert explained, "People may follow all your reviews and vote Not Helpful because they flat out don't like you. Or they may feel you can't possibly read that many books, therefore you must be cheating, and the reviews are marked Not Helpful. There's even the situations where other reviewers below you (or above you) vote Not Helpful in order to try and boost their own ranking by bringing down yours. It's vicious and pathetic behavior, but it exists."

I'm not saying that Amazon has to step in and slap the wrists of people who vote my reviews unhelpful. (Not that I'd mind....) In actual fact, they quietly do so, especially if a reviewer points out that (as has happened to me) someone visits the last 10 book reviews you entered, overnight, and marks them all down. But I've seen similar behavior on reddit (damn, there go my votes) and in other social networks. Such behavior makes it harder to draw more people into the community as active participants—a necessity for long-term survival of these services. As Duffbert said, "If you're not thick-skinned, it's best to stay out of the reviewing waters."

But it's one thing for users to squabble among themselves, under the watchful eye of a ListMom, and another for the online service itself to behave badly. It's pretty ironic, considering their businesses are inherently based on the idea of people being social.

Online communities and social networks are a conversation, even when the response is another individual voting on the post's value. By failing to effectively manage the user community—especially when they don't respond to user complaints—companies forget that they, too, are part of that conversation.