Posts about comments

After comments

Here’s my talk to Jeff Pulver’s 140Conf today on comments and interactivity, in which I argue that comments are an insult because they come only after media think they’re done creating a product, which they then allow the public to react to.

I defended comments on news sites for many years. But I think we have to move past them to true collaboration, which is more respectful and productive. There is no easy solution for civility, not identity or rating systems.

By coincidence, this appears at the same time that the New York Times publishes a story about the problems with comments, in which I suggest to the author — whose interview with me inspired my post — is often a matter of expectations: When we look at the internet as a medium, we expect it to look like media: packaged and clean. But when we realize that the internet is a place, like New York, then it’s less shocking to hear some bozo on a corner muttering “shit.”

The problem with comments isn’t them

I’m coming to think that the — or a — problem with the quality of conversation in comments online is a matter of timing:

Once we in media are finished with our work we allow the public to comment. We throw our product over the wall and let people react while we retreat into the castle and shut the gates so we cannot hear them. They know they are talking to bricks and so they shout and cover them with spray paint. Only we have the power to clean the mess but we’ve left the scene and so the castle walls are soon overrun with graffiti.

This timing — which is inherently insulting to the public — comes out of our old media worldview brought to the internet. We think the internet is a medium and that we make products for it that the public consumes.

When instead we open up to conversation earlier in our process then the conversation can become more collaborative and productive: We ask people what they know, which is a mark of respect and value. We listen to advice and requests. We end our separation from the public and join it. Waiting until we are done to listen is too late.

We must stop looking at the internet as a medium. I spent a long time this weekend talking with a reporter who’s writing about nasty comments — I’ll link to her piece when she publishes it — and I tried to convince her that the media-view we from media impose on the internet is much of the problem: When we see the internet as a medium, we expect it to be packaged and pretty, clean and controlled like newspapers and magazines and shows, and so when someone dumps a turd on that — a nasty comment — we think the whole thing is ruined, as if bad editing allowed “shit” to get into a letter printed in The New York Times.

But as Doc Searls taught me early on, the internet is not a medium — indeed, judging it as a medium brings all sorts of dangerous presumptions about control and ownership and regulation. No, Doc says, the internet is a place. It’s a park or a streetcorner where people pass and meet, talk and argue, where they are right and wrong, where they connect with each other and information and actions. It’s a public place. (And when I talk about publicness

Now judge the conversation in those terms: If you pass someone cursing on the streets of New York do you write off the place? Well, I don’t (especially because that person you pass might be me).

But all this is not to say that I accept, or we should accept the level of discourse on the internet as it is. No, I’m coming to believe that comments — which I defended when I ran sites — are an inferior form of conversation for the reasons I’ve just outlined. That’s easier to see because we’ve seen superior forms, like Twitter.

Twitter, like Facebook, is build mostly on real identities and control of relationships. I decide whom to follow and you decide whether to follow me. It’s an individual meritocracy in which each of us defines merit.

Back in the day — and still today — we hear that anonymity is the problem and that identity will solve that. That has never been the case. Identity alone isn’t enough. I may know the identity of that curser on the streets of New York but that doesn’t stop me from hearing him rant. Social controls are also needed so I can walk around him. That’s what Twitter and Facebook provide each of us. The result is better discourse. I don’t find Twitter or Facebook littered with fools and nastiness and when I do stumble upon them, I unfollow; when they occasionally spit on me, I block (if only I could instead give them their meds).

Somewhere in there is a secret to improving discourse online. Craig Newmark is talking about the need for distributed trust networks and in Twitter and Facebook I do, indeed, think we’re beginning to see the outlines of them. Clay Shirky wishes for algorithmic authority. Identity is a factor, of course. But we need to be careful about thinking that there is some system that will just clean up messy talk. That doesn’t work in life; it won’t work on the internet, which is life. What Craig and Clay are asking for is tools to help each of us have a more pleasant stroll in the streets of the internet.

But I also think we need to turn this question around and not look at the commenters but at ourselves as members of the conversation. What are we doing to improve the quality of discourse? So I return to that question of timing: When we open up and grant respect and talk with people eye-to-eye and collaborate, that creates value not just blather. In What Would Google Do? I told the story of MyStarbucksIdea.com as a platform for collaboration over conversation:

Some threads emerged from the suggestions and discussion. Many customers wanted express lines for brewed-coffee orders so they could avoid waiting behind alleged coffee aficionados with their half-this, half-that, skinny, three-pump, no-foam, Frappuwhatevers. Some customers asked to be allowed to send in their orders via iPhone. And some customers suggested—and thousands more agreed—that the chain should enable them to program their regular order into their Starbucks card so they could swipe it as they enter, placing the order and paying for it at the same time, letting them skip the cash-register line. One more proposed a pour-it-yourself corner and another asked for a delivery service. The theme—that is, the problem for Starbucks—was clear: long, slow, inefficient, irritating lines. But not one of these customers started with that complaint. Instead, they offered solutions to fix the problem. All Starbucks had to do was ask.

Should comments as a form of conversation be eliminated? No, of course not. The tool isn’t the problem (any more than blogging tools or printing presses are). If you eliminate comments that’s even more insulting than not listening to them and it risks giving up the incredible value the public can give if only they are enabled to (a value I saw so clearly in the comments under my posts here or here). The issue isn’t comments or identity or registration or tools. The issue is how you play host.