On Moderation

As you move about on the internet, on each page, ask yourself "Who moderates this?"

If it's someone's Twitter feed, that person doesn't just moderate it, they directly control it. Every tweet is either written by them or retweeted by them.

But if you click one of their tweets and Twitter shows you the responses, who moderates those? Nobody. Literally. Even the person who originally posted the tweet, unless they block the person that replied, has no control of the content shown just below to their tweet.

Contrast with a Facebook post, where at least if you made it, you can delete a comment someone else made if you find it offensive or unproductive. They get to use your post as a platform without asking permission first, but you can revoke the permission if you think it was used rudely.^[1]

So, there's a spectrum of four approaches here, which for the purposes of this article we'll call Types 1 through 4.

At one end, Type 1 systems would be something like a blog without comments (like this one) where the only way for a reader's commentary to be published would be for the author to copy it into a post and manually repost it. In short, any replies to the content require non-trivial work on the original author's part to show up alongside the original writing. This is the way Twitter worked years ago, when retweeting was a feature that Twitter clients^[2] implemented in software but was absent from the platform itself. As a part of quashing Twitter clients, they also made this a first-class feature of the platform, turning retweets (but not replies) into a Type 2 system.

A Type 2 system would be a setup where readers can contribute content back, but only after explicit approval from a moderator. This would be something like a blog where every comment must be explicitly approved, or the way retweets work today.

A Type 3 system is where the original author retains some control of the content shown in response to theirs, but only post-hoc. A Facebook post where the author can delete comments is a Type 3 system. It is a small but crucial distinction that strangers with no previous approval are automatically granted a platform directly adjacent to your own. Notably, Facebook does not have any way to disable this.^[3]

Finally, a Type 4 system would be Twitter replies where any random internet stranger can post a response that is shown just below the original content, and the original author has no control over moderating those posts.

So, as you move around the internet, apply this classification system. A Reddit thread? Type 3. A newspaper's website without comments? Type 1. (Think, "Letters to the Editor".) A newspaper's website with comments (because they drive pageviews, of course)? Type 3. A peer-reviewed, academic journal? Type 1. Comments on a YouTube video? Type 3, and yet somehow, one of the very, very worst examples of a Type 3.

A forum with very active, very strict moderators where people stick to the rules because they know they will be swiftly enforced straddles the line between Types 2 and 3, but even there you get a lot of nonsense that isn't obviously against the rules but isn't clearly productive either.

The reason all this matters is that I find Type 1 and 2 interactions to be virtually the only ones worth having. The more I cut out Type 3 and 4 from my life, the happier I am.

Now, tech companies trying to monetize eyeballs need Type 3 and 4 sites because they become a self-sustaining reaction. Running a site like Facebook is a bit like running a nuclear reactor, where you only have the moderator rods inserted just enough to keep the reaction under control, but most of the time you want it producing lots of heat to power your reactor. Their incentive is to run the site with the absolute minimum amount of moderation required to prevent the thing from melting down.

As time goes on, I think more and more that the only online places that are worth investing my scarce time and energy are ones that are moderated fairly aggressively.

And I think that this idea of having many internet communities, each run under a strict set of rules, is still compatible with the idea of freedom, but a productive, focused, deliberate freedom. You still have freedom to choose which set of rules to follow.

I think it's honestly a shame that most online communities from Facebook to Twitter to reddit are based on websites that are a decade or more old. I know there are probably many new subcultures emerging in chat services like Discord and Slack, and maybe I'm not aware of the innovations coming out of them. But otherwise we seem to have stagnated. Legacy forums with their vigilant moderators have mostly fallen out of fashion.

But there's a niche here. Reddit is the closest thing out there I think, and the ability to secede and start a new subreddit at any time contributes, I think, to the longevity of the site. (Digg used to be hailed as a similar service, but I wonder what structural differences cause it to implode when reddit didn't.)

I try to do as much of my reading now on my RSS reader (Feedly) these days, which is clearly a Type 1 system. I made myself stop reading Twitter because I found its Type 4 nature too distracting, although I've been using a service called QueryFeed to include the tweets of a few people in my RSS reader. I try to stay away from Facebook. I try to stick to Type 1 and 2 sources, and when I exhaust the available content of those (any infinite scrolling feed is a sure sign of a some kind of Type 3+ system), I make myself close the app and go do something else, like use the Kindle app to read one of the four books I have in progress at any time.

Update: I have had this post sitting in draft for a few weeks but in light of today's announcement, it seems like a good time to publish it. Twitter launches its controversial ‘Hide Replies’ feature in the US and Japan:

The addition is one of the more radical changes to Twitter to date. It puts people back in control of a conversation they’ve started by giving them the ability to hide those contributions they think are unworthy.

These replies, which may range from the irrelevant to the outright offensive, aren’t actually deleted from Twitter. They’re just put behind an extra click.

That means people who come into a conversation to cause drama, make inappropriate remarks or bully and abuse others won’t have their voices heard by the majority of the conversation’s participants. Only those who choose to view the hidden replies will see those posts.

Clearly we're still trying to figure out the right amount of moderation for healthy conversation. I don't know if this is exactly right, but the fact that they're experimenting with the formula at all, I see as a good thing.