I saw a little discussion about content warnings(external link) on Micro.blog today, and it reminded me that I have my own thoughts about content warnings – so many thoughts that I thought it’d be better to spin them out into their own post πŸ™‚ So, here we are.

Mastodon, in case you don’t know, has quite a heavily-used content warning (CW) feature. In other Fediverse software I believe the field is actually a “summary” field, so you can hide a long post behind a one-line summary and a “Read more” link. Some people use it this way on Mastodon, but others do indeed use it as a CW, warning for topics as diverse as “politics” (or often a specific country’s politics, with a tag like “auspol” or “depol” or “svpol”), “food”, “eye contact” (often abbreviated to “ec”), “mental health” (often abbreviated to “mh”) and more. There’s also a tone indicator some people add, “(+)”, “(-)” or “(~)” for positive, negative or mixed/neutral. Similarly, sometimes people use it for niche interests that they don’t think all their followers are into, like “retrocomputing” or “knitting”. Also, some people use the CW field for jokes – either hiding punchlines under the set-ups, or else just typing nonsensical CWs to “make” people open the post because it’s like clickbait – they need to know WTF that CW was on about.

Micro.blog does not have a CW feature and that’s fine with me. There’s no real need to have a “summary” field, because posts on the Micro.blog timeline already truncate after 280 characters (with a link back to the full post), or if they have a title the post on the timeline consists of just the title and links to the full post.1 I’m also not sure it’d benefit from a CW feature just for joking around with, because there’s a lot less shitposting on the Micro.blog timeline in the first place – that’s just not really the vibe of that community πŸ˜‰ So finally, we reach the question of a CW feature for content warnings, which again I’m not convinced it “needs” because if people wanna be conscientious they can give the post a descriptive title, and then people will have to click through to read the post anyway.

Manton (the developer of Micro.blog) also said in that discussion that he’s wary of adding friction to the posting process by adding CWs, which I can understand. On Mastodon I do try to CW some stuff I think might be sensitive, especially on my main account, but I don’t want to overuse the feature because for me personally I find CWs add a ton of friction to the process of actually reading my timeline. When almost every post is CW’d, I end up spamming clicks on “Read more” almost like I’m in a trance, so that on the rare occasion that something I actually don’t want to see was CW’d, I usually end up seeing it anyway. It sucks. I also find it harder to remember what I just read beyond the 3–4 second period of echoic memory, because so much of my concentration gets consumed by positioning my cursor to click more “Read more” links. I don’t want to overstate the extent of this issue, because if I slow down and re-read stuff I overcome it, but it’s not ideal.

It’s a lot better when people use their CW field as an actual summary, because then I read it properly in the first place and don’t click past it straight away to get to the “real post”. Even if people want to use it as a straight CW field, I feel like a lot of people aren’t specific enough about what their actual warning is. Like, for me, my absolute #1 “trigger”2 is animal cruelty. If people’s CW says “animal death”, most of the time it’s not cruelty but a beloved elderly pet passing away or something like that, in which case I don’t mind reading and might even want to so I can offer my condolences. On the other hand, a minority of the time it’s actually something horrible and then I really regret clicking. So that sucks too. If rather than “animal death”, the field said something like “Revelations about greyhound racing in NSW (CW animal cruelty)” that would give me much better context to decide whether or not I want to expand that post.

My personal view about content warnings is that, rather than having the entire post hidden behind a text snippet and a “Read more” link, it’s better for the post to actually include the specific term being warned about, and encourage social media users to use a filtering feature to block any content they don’t want to see. This is what people used to do on Tumblr, back in the day: they’d add tags to their post with things like “tw food” or “tw rape” or “tw suicide”3 and if those things were triggers for you, you’d configure your Tumblr enhancement extension to block posts with those tags from appearing on your dashboard. Now I think it would’ve been better if Tumblr had a) offered this feature directly and b) made it much more prominent and easily discoverable, but still, overall I feel like something like that would be a better medium. Still include a CW field, but not hide posts behind it, and make sure end users know how to filter out content they don’t want to see for themselves. This kind of filtering feature would also work well for the “niche interest” use of CWs, so people’s retrocomputing posts or whatever could just be out in the open, and readers could take it on themselves to filter out the “retrocomputing” tag if they’re not interested.4 This isn’t how Mastodon works at the moment though, and I don’t think it’s likely to evolve in that direction.

What Mastodon does have is a setting to not collapse posts behind CWs by default, but in practice that gets messy to use because of the other people who are using the field for summaries or set-ups for jokes. So IDK, I guess I’ll just live eternally in hope that people will start writing more informative one-liners.

I’ll close out this post by saying that I understand not everyone feels the same way about content warnings as me, and that’s OK, I’m not going to try convincing people that mine is the One True Way. I’ve seen enough people heaping praise on Mastodon’s CW feature that I’m guessing not everyone has this issue where clicking on “Read more” constantly takes away from their ability to actually pay attention to the posts. It does bother me a bit when I see posts admonishing others to use the CW feature more though, as if it wasn’t used enough already. Ultimately I have an easier time reading a timeline where I don’t have to specifically click on half the posts to find out what they are. It’s great to give people ways to opt out of seeing certain content they don’t want to see, but I think reader-defined filters are a better way than Mastodon’s CWs. (On the other hand, I don’t mind the CW field being used for summaries, even quick ones headline-style, or joke set-ups.) Any strategy relies on some level of cooperation between posters and readers, of course – filtering doesn’t work if posters aren’t game to include a blockable term in their posts, just like CWs don’t work if posters either don’t use them where appropriate, or aren’t descriptive enough about what they’re actually warning for. So, you know, I do my best within the confines of the systems that we have. If it were my software, though, I’d have kept the optional “summary” field labelled “summary”, added an additional field for tags like CWs or topic, and encouraged people to filter stuff they don’t like. And I would enjoy the lesser friction involving in reading my timeline πŸ™‚


  1. That goes for “top-level” posts, anyway – replies don’t truncate, although most people are pretty good about not letting their replies go on and on for hundreds of characters. Most of the time πŸ˜‰ ↩︎

  2. I put this in quotes mainly because I’ve read posts on the internet “calling out” people for appropriating the language of “triggers” from people with PTSD. I don’t have PTSD, but when I read about animal cruelty (and worse if there’s photos/videos) my heart rate skyrockets and I feel uncontrollably sad and angry and the post plays on my mind as an intrusive thought for hours if not longer. So that’s what I mean by “trigger” in this case. ↩︎

  3. As a side note, this is something else that seems to be different between Tumblr a decade ago and Mastodon now. On Mastodon I’d be surprised if a CW straight-up said “rape” or “suicide” because it’s almost like the word itself is considered potentially triggering now. ↩︎

  4. As another aside, I’ve also thought it’d be useful in some cases to be able to subscribe to specific tags of a specific user, like if I’m interested in some of what they post but not all of it (particularly if they’re the kind of person who shitposts 100 times a day but isn’t ALL shitposts). It only works if people tag their posts thoroughly, though, and because in Mastodon your tags have to be in the post body and not tacked on subtly at the end like on Tumblr, a lot of people don’t. ↩︎