Comments on community

I was an early beta tester of Newsvine, Mike’s nifty new website. I like it.

There are two ways of using Newsvine. If you want, you can simply use it as alternative to Google News, a way of catching up on the latest stories appearing on the wire. The other way of using Newsvine is to read and comment on what other users are linking to and writing about.

Personally, I find myself using the site the first way. I don’t just confine myself to the Associated Press stories though, I do also investigate intriguing things that other users are linking to. But I don’t really participate in the comments. The simple reason for this is that the comments mostly suck.

This isn’t the fault of Newsvine, it’s simply the nature of the beast. Most comments suck.

This shouldn’t be surprising. According to Theodore Sturgeon’s infamous take on the , 90% of everything is crud. One look at the music charts should be enough to confirm that this signal to noise estimate is about right.

Take Digg. It’s a nice way to find out what links people find interesting (a la but the comments attached to each link are mostly a waste of space. The more popular the link, the more useless the comments. That’s revealing. There seems to be an inverse relationship between popularity and the usefulness of accompanying comments. Slashdot also testifies to this. I’m worried that as Newsvine grows in popularity, as it inevitably will, the comments will get even worse.

So why have comments at all? In a nutshell, comments are a great way of fostering a community. But that doesn’t really answer the question; that assumes that a community is necessarily a positive thing. But is it?

Clearly, the minds behind Digg, Slashdot and Newsvine feel that the value of the 10% outweighs the ballast of the 90%. They’ve made a conscious decision that having a community built into the site is important, perhaps even the whole point of the site in the first place.

Joshua Schacter, on the other hand, made a conscious decision not to have a comment-based community built directly into A community still exists around the site; it’s still social software — it wouldn’t work unless lots of people were using it — but any benefit gained by adding comments would be accompanied by a corresponding increase in the general level of crapiness.

Clay Shirky nails it in his speech A Group Is Its Own Worst Enemy. This is a fundamental problem with social software (more so than real world situations because of The Greater Internet Fuckwad Theory). The more popular the community, the more likely it is that comments are going to suck.

Should we abandon comments completely then? Absolutely not! I’m a great believer in Tim Berners Lee’s dream of read/write web. But we should think very carefully about when and where to enable comments.

And so to blogs…

Blogs are susceptible to the same problems as social software sites (as well as having to deal with comment spamming scum). The more popular the blog, the bigger the problem. Just ask Heather or Jason.

Most blogs allow comments. There’s no doubt about it; having comments enabled is likely to increase the popularity of your blog.

But that, in and of itself, is not a good justification. It assumes that popularity is desirable. The truth is that, when it comes to personal publishing, it’s not the amount of people who visit that count, it’s who those people are why they’re visiting that’s important.

Comments are a shortcut to a Pyrrhic victory of popularity at the cost of having your pages cluttered with pointless remarks (by pointless, I don’t just mean the negative stuff: “me too!” and “great post!” achieve as little as “you suck!”). If popularity is your aim, it’s better in the long run to claw your way towards that goal on the strength of your writing or design skills.

But comments can add value. They are particularly useful on sites that have a narrow, focused scope. The focused nature of the subject matter ensures that visitors share a common interest — otherwise, they wouldn’t be there.

The more general a site’s focus, the less chance there is of it receiving quality comments. A site that covers everything from politics (Republican vs. Democrat) to computing (Mac vs. PC) is going to be flame-war central. A site that deals exclusively in the appreciation of chihuahuas stands a much better chance of forming a cohesive community.

I have three blogs. Comments are enabled on two of them. Those two are narrowly focused. One is about JavaScript and the Document Object Model, the other is about good food. While the number of comments on those sites remains low, the quality of those comments is disproportionately high. My third blog is the one you’re reading now, which doesn’t allow comments. That’s because this journal is, quite frankly, all over the place.

Sometimes I’ll write about design or code, but I’m equally likely to rant about the government or rave about a good book. My audience is, therefore, also all over the place. Geeks visit in the misguided hope that they might learn something useful but this site also functions as a way for my friends and family to keep track of what I’ve been up to (Hi, Mum!).

The inverse relationship between a blog’s diversity and popularity on the one hand, and the usefulness of its comments on the other, would seem to fly in the face of the wisdom of crowds. The two central factors in the creation of a wise crowd are that it is large and diverse. But there is another crucial factor: the individuals in the crowd should be unaware of one another’s decisions. You can be sure that the “ask the audience” section of Who wants to be a millionaire? would yield very different results if every audience member could see how every other audience member was voting.

The fact that comments are observable by default means that they effect the outcome of their own experiment. This cat of Schrodinger’s is clearly dead.

If you solicited feedback through the more private medium of email, the wisdom of the responding crowd would be greater and the quality of the feedback would increase. But that would defeat the community-building aspect: there would be no real conversation amongst your audience.

I don’t think we should be looking at comments to see conversations. It isn’t much of a conversation when the same person determines the subject matter of every dialogue. The best online conversations I’ve seen have been blog to blog: somebody posts something on their blog; somebody else feels compelled to respond on their own blog. The quality of such a response is nearly always better than a comment on the originating blog for the simple reason that people care more about what appears on their own site than on someone else’s.

The difficulty then is keeping track of these conversations. would be a good option but it relies on a certain level of techiness on the part of the responder and again, the issue of spam raises its ugly head. These days, it should be possible to replace trackback with search using third-party tools like Technorati and Google Blog Search. Expect to see that kind of functionality built in to more and more blogging tools.

I think the fundamental issue with comments is that are often enabled without reason. I wrote already about the need to justify every design decision. The same should also be true for community decisions. Does every little blog post really need to accept comments? Wouldn’t it be better to save them for special occasions?

I’d like to propose a corollary of Sturgeon’s Law for blogs:

Comments should be disabled 90% of the time.


Have you published a response to this? :