Content Warnings on Mastodon

The topic of content warnings — often abbreviated “CW” — has been coming up quite a lot over on the fediverse[1]. As a result, I’ve been thinking about content warnings more carefully than I had previously.

This is not a guide for using content warnings. I’m certainly not qualified to write that. I’m not here to tell you how you should use content warnings. This is just an explanation of how I think about content warnings and how I use them.

I’ve been on the fediverse pretty consistenly for about four years, though I would not say that I’m deeply embedded in the culture there. I’m a casual user. A lot of how I use content warnings comes from seeing how my friends use them. Some of it comes from seeing people ask for certain topics to have content warnings, which — from my privileged position — I am happy to oblige. And then some of my thinking about content warnings is the result of recent discussions and conflicts as people arrive on, or return to, the fediverse.

If you think I’ve misunderstood or missed anything important on here, please let me know. I’m @darth_mall@vis.social.

What is a “content warning”?

There is a toolbar below where you type your toots in Mastodon[2]. On that toolbar is a button labelled CW. Clicking or tapping this button will add a smaller, single-line text input above the toot area where you can enter your content warning. Toots with content warnings only display the content warning by default, followed by a Show more button that expands the toot to show the entire message.

Ok, that’s how you add a content warning, but what is a content warning? When should you use one?

This is where it starts to get a little complicated.

Outside of Mastodon, a content warning is used to signal the presence of potentially offensive or upsetting material — violence, suicide, nudity… Unsurprisingly, many people — myself included — when they first encounter content warnings in Mastodon, assume this is what they’re for. And that’s not wrong. Putting these kinds of content warnings on your toots can really help people who have been traumatized and don’t wish to be re-traumatized.

Content warnings as subjects

But it’s also quite common to use the content warnings more like a subject for the toot, like an email. The reason for using content warnings this way is that it gives people control over what they read and don’t read. For example, people will often put a content warning like “twitter” or “bird site” on toots about Twitter. This is kind because not everyone cares about what’s going on with Twitter, and the content warning makes it easier for them to just scroll right past that toot without getting sucked into reading it. During the Great Twitter Migration of 2022, I really appreciated people who put content warnings on their toots about Twitter and toots about Mastodon (usually some variation of “meta”), because there was a lot of conversation about these topics and I eventually got burned out reading them. The content warnings helped me avoid getting sucked in.

When you toot a toot — especially a toot that can be boosted — you have no way of knowing whose timeline that toot will appear in. Putting a content warning on the toot whenever you think it’s something someone might want to skip seems to me like a polite thing to do.

But content warnings are not subjects

I’ve seen one objection to the use of content warnings as subjects that, if I‘ve understood it correctly, takes issue with the conflation of subjects and content warnings. That is, a content warning’s purpose is to protect people from upsetting or triggering content; using a content warning as a subject, or as a mechanism for delivering jokes or memes, demeans true content warnings.

I can’t really disagree with this criticism. I can see how our use of content warnings for subjects and jokes might make light of content warnings for sensitive posts.

I’ve seen it suggested that there should be separate fields for subjects and for actual content warnings. I don’t know what the best way forward is, and since I’m not a maintainer, it probably doesn’t matter much what I think. For now, though, we have the CW button, and I’m going to try to use it in a way that makes everyone’s experience a little nicer.

How I use content warnings

I like to use content warnings like subjects. I also occasionally use a content warning for a joke, but for me it’s mostly subjects. Outside of using content warnings for jokes, I have three criteria I use to help me decide what topics I should put a content warning on:

  1. I see people asking for a certain topic to have a content warning
  2. I see people using content warnings for a certain topic
  3. I’m going to toot something off-topic to the local timeline on my instance

Criterion number one is just out of respect for other people. If someone is asking for a content warning, and I have no compelling reason not to use that content warning, I’ll go ahead and do that. It costs me little to add a content warning.

Criterion number two is kind of a corollary to number one. When I see people using a content warning — like “alcohol” — I assume that either they would like that topic to have a content warning, or that they’ve seen people asking for that content warning. This is how I learned to add content warnings to posts containing alcohol or photos where the subject is making eye contact.[3]

Criterion number three is about being respectful of the other people on my instance. Not all instances are topic-focused like vis.social, though, so this may or may not be relevant to you. I try to keep in mind that many of the people on vis.social are there for data visualization or art. As such, most of my posts are unlisted, because I often post just personal stuff. But occasionally I’ll post about something that might be of interest to some folks on the instance but is not strictly “on topic” — like web development. In these cases, I usually throw a content warning on there to keep the noise level on the local timeline down.

Common content warnings

These are some of the content warnings I’ve learned to use because I see people using them regularly.

alcohol

Indicates that the toot is about alcohol. Whether it contains a photo of alcohol, or just mentions alcohol, I try to put this content warning on there. Chances are there are folks on the fediverse who are recovering alcoholics. I have absolutely zero desire to make recovery any harder than it has to be by tempting them with photos and descriptions of the thing they need to steer clear of.

eye contact

This one only applies to images that have a person making eye contact. I am not an expert on this one, but I’ve seen a lot of people using it, so I do, too. My guess is that eye contact makes some people uncomfortable, and not surprising them with a big photo of someone staring at them in their timeline is just a nice thing to do.

lewd

For anything sexually explicit or suggestive. Not one I use much, but I’ve seen it around. If I do post something that falls in this category, it would get this content warning.

meta

For talking about Mastodon/fediverse on Mastodon/fediverse. Sometimes this will be a discussion of the drama of the day, sometimes it’ll be tips for using Mastodon. There can be a lot of meta discussion on the fediverse, and some people are just not into it, so it makes the fediverse a little nicer for them to content warn this stuff.

mh

Stands for mental health. When people are posting about their own mental health, they tend to put this content warning on it. It’s often accompanied by either a “+” (plus) or “-” (minus), to indicate the tone of the post (more on that in a moment).

pol/uspol

Stands for politics. You’ll often see this one prefixed with a two-character country code like “us” or “uk”. Not everyone wants to get sucked into the latest political drama, this helps them avoid the outrage cycle if they’re not up for it. I always slap one of these on my political posts, even if it’s something seemingly innocuous like “I voted!”

shitpost

I recently started labelling my shitposts (Urban Dictionary) with a content warning. For me, this serves two purposes:

  1. Anyone who doesn’t care to read whatever random garbage thought I just shared can easily skip over it
  2. It signals to anyone who does read it that it’s not to be taken seriously, so they don’t need to reply to engage me in a debate

Win-win.

twitter/bird site

Toots about Twitter (nicknamed “the bird site” on the fediverse). Much like “meta” posts, not everyone wants to talk about Twitter.

+/-/~

The “+” (plus), “-” (minus), and “~” (tilde) symbols are used in conjunction with other content warnings to indicate the tenor of the toot. A plus is a positive toot, minus is negative, and tilde is either a little of both or a neutral toot. You might see something like “mh, +” in a content warning, which would indicate that someone is talking about their mental health and that the toot is positive in nature. Some days you might be up for a “mh, +” toot, but not a “mh, -” toot. Some days you might just not be up for a “mh” toot at all.

Other obviously upsetting topics

There are some topics — the kinds of things you think of a traditional content warning being used for — that are pretty obviously potentially upsetting: violence, racism, suicide, death. I always put a content warning on these kinds of posts when they come up. Sometimes I’ll be more specific, like “gun violence” or “shooting”, instead of just “violence”.

Asking for content warnings

I am careful about how I ask for content warnings. Again, I am not an authority on these things, so I can only tell you the guidelines I use when asking for content warnings.

For example, I might toot something like:

Hey folks, I’d really appreciate content warnings on World Cup toots. Thanks.

I haven’t addressed any individual specifically. I haven’t told anyone what they should do. I’ve just expressed my desire to have content warnings on World Cup toots.

Here’s what I would not do:

@arsenalfan2000 would you mind putting a content warning on you World Cup posts? Thanks.

The difference here is that I’ve specifically asked someone to use content warnings the way I want. Even though the toot is polite, and the World Cup is (probably) not a sensitive issue, I still wouldn’t do this. Let’s get into why…

Policing content warnings

This has been one of the big conflicts around content warnings recently. I have to be clear here that I cannot speak from experience about this matter. As a straight, white, cisgender man, these issues do not impact me directly, so I can only convey my understanding of these issues as they have been explained by the people who actually are affected.

People who have been or are being oppressed — in particular I’ve seen a lot of Black of people discussing this issue — choose not to use content warnings when they are talking about racism or violence in their lives and their experiences. This is intentional. Asking someone — or worse, telling someone — speaking about their own experiences to hide their lives behind a content warning is silencing. If you — like me — are white, this is even worse. There is a long history of white people trying to sweep the awful truths of racism under the rug using civility as an excuse. I’d be willing to bet the same is true if you are straight and talking to queer folk; or if you are cis and talking to trans folk; or if you are male and talking to someone who isn’t male.

So asking or telling someone to put a content warning on their toots is something I wouldn’t do. It could easily be seen as trying to silence marginalized communities. My approach here may be overly conservative, but given the amount of privilege I enjoy, I prefer to be extra cautious in this regard. If someone’s lack of content warnings poses a problem for me, I might unfollow them, add filters for certain keywords, mute them, or even block them.

How you should use content warnings

I can’t really say. I tend to think of using content warnings as a form of good manners. You show respect for other people’s feelings and well-being by using content warnings; giving them agency over what they are exposed to on the fediverse. But there are times when good manners aren’t called for. You have to decide for yourself whether or not you think your toots should have content warnings or not.

As a member of the most privileged demographic on the planet, I am sensitive to the requests for content warnings from others. Personally, I will continue to listen to what others have to say about how we can make the fediverse more welcoming, and my use of content warnings will probably change accordingly.

Again, if you think I’ve misunderstood or missed anything important on here, please let me know. I’m @darth_mall@vis.social.


  1. I’m talking mainly about Mastodon here, because that’s what I use, but it is more accurate to say “the fediverse” because Mastodon is really just an ActivityPub client. The collection of servers running software that communicates over ActivityPub is called the fediverse. ↩︎

  2. I only use the official Mastodon web app, so I can’t speak to where the buttons may be in other clients, or how they’re labelled. ↩︎

  3. One thing I’ve learned being on the fediverse for four years and seeing all of the content warnings is that my understanding of what might be potentially upsetting was too narrow. Topics that seem innocuous to me, can be quite upsetting to others for a variety of reasons. ↩︎