Four Ways to Improve the Culture of Commenting

In my Riff for this past weekend’s magazine, I write about the history of online comments and how they came to often take such an off-putting form. Here are some things we could adopt to help build a better commenting environment:

User-driven moderating: Create more robust systems of user moderation and teach people how you want them to participate online by training and rewarding them. The best argument for this comes from economics, specifically the work of the Nobel Prize-winning economist Elinor Ostrom, who showed that commonly held resources (like fishing grounds or groundwater) are more sustainably managed by communities of users, instead of by government agencies or the marketplace.

I often find that the problem with comments isn’t that they’re offensive or off-topic; it’s that they’ve distracted me with irrelevant information or told me what I already knew. (Yes, some people don’t like Obama.) I wish I could flag a comment or even a whole thread as a waste of time for other readers. The tragedy of the comments is a tragedy of the commons, because the unreplenishable resource that has been overexploited when comment threads go awry is the finite amount of attention that we have to spend reading. Any individual commenter is simply doing what the system encourages him to do; in aggregate, however, all of these decisions eat away at the attention.

“You have time to create something beautiful. You have time to read the comments sections. You do not have time to read both,” Shane Liesegang tweeted (@avoidcomments). I’m not comfortable with eschewing comments altogether, because I assume that the expertise that finds its way into bylined articles in the newspaper is a small subset of the expertise that exists in the world. If the comments section enables a larger amount of the expertise to reach a wider audience, then I’m all for it.

There’s still a problem, though: How do you know where to go for comments that are worth reading? That leads to the next improvement:

Commenting weather systems: One of the tools that’s used to analyze Twitter and other social media is called sentiment analysis. This is a machine-based method of measuring the emotional dimensions of what people say. Computers analyze huge chunks of language that people use when talking about products, institutions, events and other things, and then they extract how people feel, either negatively or positively, about those things. Someone should create an Internet-wide sentiment-analysis system that would track, in real time, the commenting “weather” on various participating Web sites. Positively oriented threads are good weather and negatively oriented ones are inclement, however “positive” and “negative” are defined. Publications could even provide a view of the commenting weather within their own sites. It’s in their interest to do this sort of thing, especially if they’re investing in human moderation and filtering algorithms. If those tools are improving the level of discourse but no one knows that, what’s the point? This could also lead to a culture of rhetorical storm-chasers, people who flock to virulent comment threads and report back to calmer parts of the world.

A commenting weather system would promote another development:

Connoisseurship: Increasingly, when people recommend an article to their friends, they also note the quality of the comments, and indicate that a particular flavor of a comment thread is the reason to go to — or avoid — related content. You might have already seen this happening informally. I see it all the time on Facebook.

This sort of connoisseurship should be encouraged to grow around the culture of commenting, and it should start with awards: a Pulitzer Prize or something similarly prestigious given to comment threads, with the award shared equally among the commenters, the moderators, the writer and the publication. The entire content + comments, as well as a moderation log, would be submitted as a package, so that judges could see the quality of the article, the value added by the comments and the decisions that moderators made about curating the comments.

In order to foster connoisseurship, another tool is needed:

Searchable comments: There’s no better indication of the hierarchy of content on the Web than the fact that comment threads are generally not searchable. For this blog post, I wanted to mention a remarkable comment that accompanied a photo-driven story in Boston Magazine about the capture of the marathon bomber, but I didn’t have time to search one by one through the more than 3,300 comments. Similarly, I’d love to credit to the person who wrote the pithy statement I quoted in my Riff (“the comments are where the real America is”), but I couldn’t find it again.

Would any of these things change comments? Maybe not all of them. But if the Web is about participation, we could enlist more of the off-line world’s tools to reward good participation — not just gripe about the bad actors.