We should ask ourselves what value on-article comment sections provide. I believe HN is good because it serves a specific trade purpose and caters to a specific

We should ask ourselves what value on-article comment sections provide. I believe HN is good because it serves a specific trade purpose and caters to a specific high-end niche audience. HN's participants are willing to cross a significant participation threshold to be part of the community here. I believe that kind of structure needs to be in place to get good community participation. I guess I'd summarize it with these points: 1. require significant investment from the userbase to participate successfully 2. promise to assist the userbase with something that is critically valuable to them in exchange 3. regularly deliver on 2 to keep 1 worthwhile 4. provide reputation tracking and management utilities so that the user can cultivate a profile that reflects the investments made in point 1 5. for recurring participation, provide variable rewards that trigger the brain's hooks for surprise, which translates to enjoyment. HN hits those points, but blogs definitely don't and don't want to. They want to bring in barely-interested readers from search, from anywhere on the web. Many of these readers won't even make it into the comment section. Thus, a good community tacked on to the bottom of their articles seems unlikely. If those principles are required to cultivate a worthwhile community, the community should always occur external to the publication of the article. The community needs to be the centerpiece, not the article. I use HN this way; the discussion is the primary thing, the articles are the subjects submitted for the community's discussion. The other caveat is that it's difficult to provide points 3 and 5 when you're just starting out. From what I've seen, it practically always has to be artificial until the momentum becomes self-driving (if there is a physical community that uses the online forum for spillover, this may not be applicable; this is basically what happened with HN). We need better solutions there. Since individual blog posts have certain quantities of Google juice, comment sections will be overrun with spa--err, "SEO professionals". Other participants are often low-effort drive-bys. If the community isn't the principal focus, participation will be spotty and it will be hard to develop elements fundamental to meaningful community engagement. Facebook and Twitter are normal people making normal comments on random stuff they see. These people generally feel a compulsion to let their feelings out and Facebook/Twitter provide it. I believe this is probably what was originally intended for blog comments, but because Facebook/Twitter are real, stable communities, the participants are inclined to leave comments there instead of on the target article itself. These comments are often loose and instinctive, which is not necessarily to say they're invalid or worthless, but a community won't form around individual postings because there's no common unifying dictum (point 2 is unfulfilled, and point 1 is minimal on random FB/Twitter posts).