Today I saw a tweet about danah boyd's speaking experience at Web Expo 2.0 earlier this month. Intrigued I followed the link and read her account of having a (twitter) back channel posted behind her during her keynote presentation. I intially clicked through because the tweet included a pretty sensational quote from the blog post which said danah felt objectified by the experience. Lots of other people have blogged this story (and I'm sure there were others too). What happened was pretty nasty. Please do read about it and think about how you would have responded in that situation. Write up your own response if you want. Then when you've finished, come back here. What I want to talk about here is a technical solution to the technical problem that I saw described. If we were to look only at the social part of this problem, the solution seems very obvious to me. It is the job of a conference organizer to ensure that participants enhance, and do not detract from, the experience of the conference. In the simple case of bad behaviour, this is how conference organizers should deal with the situation:
- Decide ahead of time what is considered appropriate behaviour for the live event. You may want to use a variation on the Wiscon or Peguicon Rules, or the Ubuntu Community Code of Conduct.
- Decide what the punishment will be for breaking the rules.
- Publish the rules and the punishments on the conference Web site. Include an "I have read the rules" statement as part of garnering a ticket to the event.
- During the event conference organizers must enforce their code of conduct. This includes actions that happen in meatspace (i.e. in person) and also the digital space.
But this only covers (1) the establishment of a code of conduct and (2) how to react in the case of bad behaviour. We wait for the participants to screw up, and then we punish them. The organizers of Web 2.0 took down the twitter feed during danah's presentation. And then put it back up again. As far as I can tell no action was taken against individual commenters for inappropriate comments. Perhaps none of the individual comments broke the code of conduct and were simply "too noisy." Noise isn't necessarily a punishable offense. It's just rude. Generally speaking the audience actually wants to hear what the speaker has to say. We pay to see a movie; we pay to see a speaker. We do not pay to listen to the people next to us complain about how bad they think the movie is.
Conferences are using Twitter streams (and/or Status/Identi.ca streams). The streams are (generally) not moderated. And when posted during the presentation, behind the presenter, they have the ability to make the audience more powerful than the presenter. Blah blah blah democratization blah blah blah. However, when comments are made in the human fashion (by people yelling or whispering or otherwise using their body to communicate) others can respond in a human way. When a presenter can engage with the audience, they can perform their own crowd control. But when the comments are made in the digital space, such as in danah's case where the comments were posted behind her back, it is impossible to use the "normal" techniques. The digital space also lingers. If I yell out, "you suck." I've said it, and then it's gone. But if I tweet it, my comment remains in the digital sphere for a much longer period of time.
This is where the technology gets interesting. Currently conference organizers who use back channel (as in the Web 2.0 Expo example) choose to allow all comments to linger...loudly...behind the speaker. The only way to clear them is for someone else to add their comment to the space, or for the refresh period to wipe the comment out. No matter how fast the refresh period it will still be longer than what it would have been were the comment shouted out. This is why written comments are generally pre-screened for television/radio/broadcast. If there is no ability to interact with the comment generator in real time, comments get screen before being published. (Aside: I used to be one of these screeners when I worked at TVOntario.)
Our current tools force a conference organizer to publish everything that is said with a specific hashtag to the screen. (Let's ignore the part where I don't like a feed being published BEHIND a presenter DURING their presentation. If you want to engage in back channel, it should be your choice...if you left your laptop off you should not be forced to read comments while trying to focus on a preseter. That was a long bracket even though I was trying to ignore that part of it.) So let's pretend you're a conference organizer who WANTS to publish this feed, and the presenter has agreed to it. Many conferences have rules on what is considered acceptable behaviour (including the consequences for breaking the rules). Generally this applies to things happening in the human space, not the digital space. Generally censorship in the digital space is considered bad. But it is ultimately the conference organizer who is publishing the comments. The comments don't happen in meatspace, and yet they are happening behind someone who is presenting in meatspace. (Another aside: I do think that behind is better because it does allow the presenter to deliver their presentation without also having to scan the comments on top of delivering the presentation. It would be appropriate, however, during a Q&A session to have the presenter read the comments at the same time.)
Since this is a technology-induced problem. A technological solution is appropriate. What if we were able to moderate comments BEFORE they were published to the screen? What someone could make a human judgement before each twent (a new typo for tweet + dent) was posted. What if someone could say to every comment, "enhance" or "detract" and publish accordingly? Although it could be the audience who makes that decision (a la backchan.nl), I'd rather the audience be mostly focused on the speaker, not audience comments. Having community moderation also puts even more of a lag between the comment and its context. A single human could process the stream faster than a community could moderate it. The Status/Identi.ca code base does have the ability to moderate comments post-publishing. But what if a new feature was added which allowed a Group Moderator/Owner to select which twents (this word may stick, sorry about that) were posted for public consumption.
When a conference organizer chooses to stream content, they take on the responsibility of being the publisher of all content that stream generates. Conference organizers should have the ability to moderate what content is published. (They may also choose not to moderate, but they should have the technical freedom to make that decision.)
If you were to design the UI for conference (or meatspace) back channel moderation, what would it look like?