This afternoon, several of our readers questioned our decision to close the comment thread on Jaron Lanier's post about WikiLeaks, "The Hazards of Nerd Supremacy." The discussion transformed into an extended Twitter conversation with some of my favorite writers, professors, and readers about the ethics and strategy of that decision. I'd like to walk through my thinking with you all here.

Before we dive into the details, I want to tell you where I'm coming from, philosophically and strategically.

First, I love Atlantic readers. This isn't pandering; I go home every night thankful for the audience that we have because I know that I've got it good. But we're a totally open website and we can draw trolls just like anywhere else. I kinda love our homegrown trolls, but I have no patience for people who visit once just to say something mean. They demean the community.

Second, I am not religious about the necessity of comments. Comments have a history just like every other literary genre and that entails conventions. Many of the general conventions around Internet discourse developed in ugly ways. (That's why oddball commenters are so delightful, like this limerick-writing guy.) Commenting standards differ across the Internet -- our unwritten code tends to require basic civility -- but when you have an article get widely distributed, the standards of the local site tend to collapse from the onslaught of outsiders. I also firmly believe that comments are not always the best conduit for critique, feedback, and reflection. Theoretically, at least, one could design a better general system, or one that could be used in some special cases.

Third, a closed comment thread does not prevent the propagation of critical responses in today's world. To suggest so is to ignore the link. In age where blogging -- micro, meso, and normal -- could not be easier, most discussion of a story takes place off-site anyway. In fact, that's one problem for us sites that would like to capture the conversation-creating power of a story.

Ok. Now, to the specifics.

Late last week, Jaron got in touch with The Atlantic about a piece that he had written for the German magazine Focus, but which did not have an English-language home. After a couple of emails and a quick conversation, we agreed to take the story. Jaron's book has been more influential than his detractors would like, and even though I didn't agree with most of what he had to say, I think of the tech channel as a place that can host alternative viewpoints. That is to say, I feel OK posting things with which I disagree. (It might sound silly to people who don't work here, but the tolerance for difference that was baked into our magazine 153 years ago is still important to us. I mean, where else could Megan McArdle and James Fallows work in adjacent offices?)

In any case, while we were talking about the piece, I told Jaron that I wanted to post my own response to it. Jaron made the request that I hold off on the response for a short while. Because his essays are long and occasionally difficult, he worried people wouldn't read the actual essay, instead jumping down to the comments or over to my response. Given Jaron's long history and strong positions on some hotly contested issues, it did not seem unreasonable that his piece would be marred by nasty trolling instead of real engagement with his ideas and arguments.

This is a tricky issue because I want to do right by our writers and readers. (After all, dozens of our readers *are* our writers.) In a competitive market for ideas, part of what we can offer to writers is a little more control over how their work might appear. The context that surrounds a piece just keeps growing. Just a few years ago, media organizations held complete control over the presentation of work. I'm glad that there is much more flexibility now, but -- as someone who writes for the interwebs every day -- I don't think it's unreasonable to want to control *how* your work is read on its original publication. Of course, we have to balance the desires of these producers with the expectations of readers.

In this case, I agreed to a closed-comment policy. I thought many of those who would want to respond to Lanier would use their own blogs, Twitter feeds, and other venues to wrangle with his work. My plan was to collect up those responses and post them to the site -- and then link that piece in with his original essay. I also wanted to respond myself, made a call for responses on Twitter, and directly reached out to two people I thought might be able to provide good counter essays (Biella Coleman and Zeynep Tufekci).

And, I figured, we'd never closed comments, so it seemed worthwhile to see what would happen if we did. The Technology Channel is less than four months old. *Everything* is still an experiment. Hell, on The Atlantic, we have thriving communities that make extensive use of comments (Coates, McArdle) and that use alternative mechanisms (Sullivan, Fallows). I think what I do is a lot more like the former, but perhaps we could deploy the latter in some circumstances?

So, I clicked the little checkbox in Moveable Type to not accept comments and went on with my business for the day, watching the feedback roll across Twitter. Later that evening, my associate editor Nick informed me that comments were in fact rolling in. That's because Disqus, our commenting system, is not actually linked with the Moveable Type check box.

That was not good. In the time between posting and when Nick noticed the problem, 10 comments had been posted. I hadn't held up my end of the deal with Jaron -- and I realized it would seem as if I was shutting down comments after they turned nasty (which they mostly had).

At that moment, it made sense to me to explain the situation reasonably and shut down the comments, providing my email address as an alternative venue for expressing feedback about the piece. It was not a perfect solution for Jaron, nor for readers, but it seemed the best that could be done.

This afternoon, I found myself on the hot seat with a variety of people that I know and respect for that decision. Three early queries came from  NYU journalism prof Jay Rosen and Patrick LaForge of the New York Times. Not exactly the people you want to end up on the other side of a Twitter debate about online communities with.

"If you read the Jaron Lanier essay in The Atlantic, what do you suppose is all about?" Rosen asked. His link went to my explanation for closing comments. LaForge followed up, saying, "Odd how Atlantic has shut down comments responding to this WikiLeaks/Anonymous piece." Soon, Kathy Gill, who teaches digital journalism at the University of Washington, had joined the discussion, as had Mathew Ingram of GigaOm, who noted that "it's good to have other avenues *in addition to* comments, not necessarily instead of."

When I explained the situation as I have in this post, Gill and Rosen were not exactly satisfied. (Really, most probably were not.) Rosen made the very fair criticism that Jaron hadn't linked to or quoted the nerd supremacists he opposed. As his editor on the piece, that criticism falls to me. He's right. There should have been a lot more links in the post. (Some other time I will attempt to describe how editing relationships are also growing harder to manage these days.)

Gill remained concerned about the lack of comments specifically. "Email does not create good *public* discourse; you describe letters to the editor model," she wrote. Corinne Marasco joined the conversation around this point, too. "If Lanier didn't want comments then why is he writing online?" she asked. She also pondered The Atlantic's heterogeneous approach to commenting.

As I responded (and parried and dodged), I realized that the experiment of turning off comments had gone, as I told Gill, "quite badly." I think the honest mistake of accidentally not turning off comments and then having to do it after a few hours compounded the problem, but was not fundamental to this situation.

The truth is that The Atlantic Technology Channel's strongest supporters and readers have come to expect that their community allows commenting. And if we were going to turn off comments for any reason, A) it better be a damn good reason and B) we better clearly indicate other ways of interacting both with each other, the editors, and the author. I wasn't clear enough about what was going on at the outset nor when I actually shut off comments after a few hours.

I don't want to rule out ever turning off comments again, but I do know that we'd execute very differently. Oddly, I'm heartened that we've developed enough of a reputation as an open and good place to talk about technology that the inability to interact on the site is perceived as an "epic fail," as one reader told me. We are a community now; certain rules have emerged.

And here's the other lesson I learned, which may be more generalizable. I'm an experimenter and so are many of the staffers here at The Atlantic. We've been tremendously lucky that most of the things we've tried have worked. But you don't always experiment for the good times. You need to have things not work sometimes. There's nothing like a (very) public learning experience to focus the mind on the things that matter for your site.

Oh, and expect those responses to Lanier's piece up soon. I promise comments will be open.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.