Content warnings on Mastodon

Content Warnings on Mastodon

The Potential and Drawbacks of the ‘Friendliest’ Social Network.

A post by Jacqueline Burggraf, Laura Gil and Anna Kuschezki

Wednesday, 18 January 2023

Media, Technology and Society

Elon Musk’s acquisition of Twitter in October last year triggered an exodus to other social media platforms. Not only do people object to Musk’s frequent controversial statements and opinions, they are also dissatisfied with the way he is running the platform and transforming Twitter as its new CEO.

Mastodon is considered the social media platform most similar to Twitter – which is why it has seen tremendous growth in user numbers over the past few months. On 26 October 2022, the day before the Twitter acquisition was finalised, there were 5.6 million accounts on Mastodon. By 14 December, this figure had risen to 8.4 million accounts – in less than two months.

Whilst Twitter and Mastodon share many features, there is one feature on Mastodon that is unfamiliar to former Twitter users and has sparked a great deal of debate: content warnings.

Let’s take a step back and explain:

What are content – or trigger – warnings?

Let us first look at the term ‘trigger’ from a psychological perspective. Here, the term ‘trigger’ refers to certain external stimuli that can trigger negative experiences in people with mental health conditions. Most often, the term ‘trigger’ is used in the context of post-traumatic stress disorder. In this context, triggers are linked to or resemble the traumatic experience. These triggers can be specific people, situations or even sounds. For people affected by PTSD, this means that they relive traumatic experiences in the form of what are known as flashbacks. These flashbacks are an involuntary, uncontrolled and sudden re-experiencing of a traumatic situation and place a great deal of stress on those affected, especially if they are unable to prepare for the trigger. Based on the trigger warnings, everyone can then decide for themselves whether they wish to engage with the content or would prefer not to. Thus, the benefit of trigger warnings is to protect people from potential triggers.

However, content warnings originated in forums for victims of sexual violence. There, their purpose was to prepare readers for explicit content. Trigger warnings are similar to other content warnings in that they all aim to protect users. The difference, however, is that trigger warnings apply only to people suffering from PTSD. They are similar to the Voluntary Self-Regulation of the Film Industry (FSK), which regulates certain content for children and teenagers. The difference is that trigger warnings are aimed at adults. In recent years, trigger warnings have found their way into commercial media. There, they are increasingly used for content other than sexual violence and are intended to protect people from disturbing content, whether or not they suffer from PTSD.

What are the reasons for using content warnings?

Many educators advocate for content warnings when they are used to protect people with PTSD or other serious mental health conditions. Additionally, content warnings can be used as a means of inclusion to help people with mental health conditions participate in classes at school or university, for example.

Essentially, content warnings are designed to prevent people from being exposed to topics that could trigger flashbacks or other mental health issues. The warning thereby enables these individuals to prepare for the topic and thus avoid the trauma response.

Another reason for content warnings can also be the so-called Werther Effect. The Werther Effect demonstrates that the type and style of reporting on suicide can influence society’s perception of suicide. Consequently, a certain type and style of reporting can lead to an increase in suicide rates.

Is there also criticism of content warnings?

Yes, there is!

Experts fear that trigger warnings could be overused and thus lose their meaning. Furthermore, the increased use of trigger warnings for topics that do not require them could lead to the trivialisation of PTSD, and sufferers might feel patronised as a result. According to critics, the fact that such trivialisation is already taking place is demonstrated by the fact that the term ‘trigger’ has entered the everyday vocabulary of many people and now describes a wide variety of different things and situations, even independently of traumatic experiences. Another criticism is that the frequent use of trigger warnings makes the term ‘trauma’ increasingly imprecise. It is no longer about trauma in the psychological sense of a one-off, severe and overwhelming experience that is relived upon being confronted with triggers, but rather about these triggers being seen as causes of traumatisation.

Now back to Mastodon:

Mastodon was created in 2016 by the Russian-born German software developer Eugen Rochko. According to Rochko himself, he created Mastodon because he used to love Twitter but became dissatisfied with the platform. Mastodon’s mascot is – you guessed it – a mastodon. Mastodon is based on open-source software, which means that the social media network is entirely self-hosted. As Mastodon is a non-profit organisation, there are also no adverts on the platform: to this day, everything is funded through crowdfunding. So how does Mastodon compare to Twitter?

Mastodon’s mascot is – fittingly – a mastodon. (shorturl.at/jwFPX)

First of all, both platforms can be classified as microblogging services, where people can post short messages for their followers. Users have an account with a profile picture, a bio and a username – visually, Twitter and Mastodon are very similar. The main difference is that Twitter is centralised and Mastodon is decentralised. This means that Twitter is run on a single server by one company, whilst Mastodon consists of a collection of many servers – or ‘instances’, as they are known on Mastodon – which form a large network. Each instance is run by its respective owner, who can set their own rules and determine the content for it. An instance can be as small as a single account, but can also have thousands of members.

Users can follow other individual accounts on Mastodon, just as we are used to on other social media platforms, but they can also join the instances mentioned above. There are many instances on Mastodon, covering various topics, interests and groups. The largest instance is mastodon.social, which is run by Eugen Rochko himself. Mastodon also has timelines, like most social media networks. Users have one federated timeline, where all their recent posts are gathered, and can have multiple local timelines that are limited to posts from just one instance. Users can write posts, reply to posts, and ‘boost’ posts – which is the equivalent of retweeting on Twitter. However, there is as yet no Mastodon equivalent for ‘quote-tweeting’, which involves retweeting a post whilst simultaneously leaving a comment on it. Many recent Mastodon users, particularly those migrating from Twitter, have criticised the lack of this feature.

Content warnings as a feature are unique to Mastodon, at least when considering the major social networks. Whilst many people use content warnings on Twitter, Instagram and other platforms, they can only include them in their captions or post texts themselves – and the social network itself is not involved. In comparison:

Here’s how to use content warnings on Mastodon

When posting, users can add a content warning at the top in a separate field if they choose, and they can also add a reason for the warning or another short phrase stating that they are using a warning. From the viewer’s perspective, when a content warning is present, the rest of the content is hidden and only the warning is displayed. Users must actively click ‘show more’ to view the full post.

Mastodon also has a content warning solution for images. When users add media to their posts, they can mark the content as ‘sensitive’, which causes the image or video to be hidden behind a blurred overlay when the post is published. Again, users must actively click on the image to view it unblurred. When users add a content warning to a post and include media, Mastodon will automatically blur the media thumbnail as well. This means that users can use the image blurring feature without a content warning at the top of the post, but not the other way round.

The content warning feature on Mastodon: From the poster’s perspective, with the post collapsed, and the post visible. (Source: shorturl.at/HIQVW)

The concept of content warnings has been discussed many times by experts and online. There are many people who do not see the need for content warnings on social media, or even view them as harmful.

So why does Mastodon offer this feature?

Mastodon and its creator, Eugen Rochko, have a self-proclaimed ethos of being friendly, polite, curious, open, caring and decent. This means everyone should feel welcome and be able to use the platform safely. In Mastodon’s own online guide on how to use the platform, they suggest using content warnings to flag triggering or offensive content – but since anyone can freely add and word content warnings in their posts as they wish, without any restrictions, the warnings are used for a wide variety of content.

Is Mastodon really that friendly?

As mentioned, Mastodon is generally considered to be a very friendly platform. Researchers wanted to find out whether this impression is accurate and examined 13,590 trending posts in 2022. The posts were analysed to determine whether they conveyed a positive, neutral or negative sentiment. Al-khateeb et al. found that most of the posts tended to be negative. However, the tone of a post alone does not reveal much about how people interact with one another. Therefore, in a further step, the posts were analysed for toxicity, and the researchers concluded that only 4 per cent of the posts exhibited toxic behaviour – the platform appears to have earned its reputation for friendliness.

However, the platform’s decentralised nature can cause some issues here, as there are many different communities across various servers. Each server can set its own rules and ban different topics. Another problem is that every server has its own moderators, so dealing with hate speech, for example, can be very difficult.

Which content warnings are used and why?

Among the most common content warnings used on Mastodon are eye contact, alcohol, and mh (mental health).

The table above shows common warnings that are frequently used on Mastodon but not seen as often on other platforms. However, there are, of course, also the more common warnings relating to racism, violence or suicide. Another interesting point is that many people use the feature to give their posts a headline or short summary.

This shows us that warnings on Mastodon are actually used more frequently than we might be used to. To understand why this has developed on Mastodon, we should look at the groups of people from which the platform emerged: for a long time, Mastodon was a niche platform where marginalised groups could exchange information. These people often switched to Mastodon to be able to connect with their community without facing hostility or insults. Over time, they have created a safe space where mental health issues are accepted and interactions are generally respectful. A particularly interesting aspect of how warnings are handled on Mastodon is the concept of consent, as you have to actively expand posts and agree to read them.

Criticism and possible solutions

With the many people leaving Twitter to join Mastodon comes a wave of criticism regarding the use of content warnings – particularly as these users are new to the way warnings are handled on Mastodon. One criticism frequently raised is that, because there are so many warnings, those concerning truly sensitive topics lose their impact. Another issue is the clarity of the timeline. If all posts contain warnings, people have to decide on a case-by-case basis whether they want to read each post and then expand it. Many people also complain that this obscures important topics. If all political topics are flagged with warnings, for example, they may not be discussed sufficiently.

However, solutions already exist for most of these issues. Some solutions have already been implemented or have been available for a very long time. For example, there has long been a solution to the criticism that it would be difficult to open each post individually. In the feed, users can choose to view all posts without warnings. For people who feel uncomfortable with certain topics, there is the option to filter hashtags and words. Posts containing these will then not be displayed. Furthermore, you can, of course, also mute certain users, unfollow them or block them.

Then there are some proposed solutions that are currently under discussion. For example, many users would like to see the ‘Content Warnings’ feature renamed to ‘Content Notice’. This would mean that new users would not be taken by surprise by how the feature is used, and the name would be more in line with its practical application. There is also a discussion about whether a different feature should be used. The argument is that there could be one section for content warnings and a separate section for headlines.

So what does all this tell us?

Content warnings are actually used very frequently on Mastodon, even for topics that might seem harmless at first. However, if you take a closer look at many of these warnings, you will notice that they are quite helpful for certain people. Of course, there are also warnings that, objectively speaking, do not make sense. For example, when the feature is used to convey jokes, such as a ‘knock-knock’ joke.

However, you get the impression that long-standing users really take the warnings seriously and are very respectful towards one another. You can also see that it is a well-established culture. Many users include warnings because they see them on other users’ posts – and those other users explain why a warning is necessary.

Overall, the warnings are used as experts would recommend: the word ‘trigger’ is not mentioned, extremely alarming words are not used, but the topic of the post is stated briefly and factually, and each person is then free to decide whether they want to see the post or not.

What remains unanswered, however, is whether trigger warnings obscure important issues too much. In the near future, we will see how Mastodon develops and whether it can become a replacement for Twitter in all the ways users expect and hope it will be.

Sources

Al-khateeb, S. (2022). Dapping into the Fediverse: Analysing What’s Trending on Mastodon Social. In R. Thomson, C. Dancy, & A. Pyke (Eds.), Social, Cultural, and Behavioural Modelling (pp. 101–110). Springer International Publishing. https://doi.org/10.1007/978-3-031-17114-7_10

Brunner, M. (2019). Trigger warnings: On the politicisation of a trauma therapy concept. In C. Busch, S. Elsuni, J. Erkens, L. Fischer, L. Gorelik, D. Krieg, S. Lohaus, S. Maani, & H. Haruna-Oelker, Trigger Warnings: Identity Politics between Defence, Isolation and Alliances. Verbrecher Verlag. https://public.ebookcentral.proquest.com/choice/publicfullrecord.aspx?p=5820241

Cantrell-Kringle, M. (@mckra1g@mastodon.social). (18 November 2022). Selfie, eye contact. It’s cold and windy, which means I slather my face in powdered doughnuts before I leave the house. But fr… [Mastodon post]. Mastodon. mastodon.social/@mckra1g/109366817126745030

Columbo, m. (16 August 2018). A non-computer-person’s guide to how Mastodon instances work. Medium. Retrieved on 14 December 2022. https://medium.com/@jimpjorps/a-non-computer-people-guide-to-how-mastodon-instances-work-da6ceac1994a

DeGPT. (n.d.). Post-traumatic stress disorder. Retrieved 16 August 2022, from www.degpt.de/informationen/for-those-affected/trauma-and-consequences-of-trauma/how-to-recognise-trauma-related-disorders/post-traumatic-stress-disorder

Jarvis, J. (21 November 2022). On Joining Mastodon. Medium. Retrieved on 14 December 2022. https://medium.com/whither-news/on-joining-mastodon-d539eed5e41a

Jones, P. J., Bellet, B. W., & McNally, R. J. (2020). Helping or Harming? The Effect of Trigger Warnings on Individuals With Trauma Histories. Clinical Psychological Science, 8(5), 905–917. https://doi.org/10.1177/2167702620921341

leahoswald. (8 November 2022). Rename content warning (CW) to content notice (CN) · Issue #20117 · mastodon/mastodon [Suggestion]. GitHub. github.com/mastodon/mastodon/issues/20117

Mastodon gGmbH. (n.d., a). We develop Mastodon. Retrieved on 14 December 2022. https://joinmastodon.org/about

Mastodon gGmbH. (n.d., b). Posting to your profile. Retrieved on 14 December 2022. https://docs.joinmastodon.org/user/posting/#cw

Mastodon Users [@mastodonusercount@bitcoinhackers.org]. (14 December 2022). 8,426,871 accounts +950 in the last hour +24,003 in the last day +231,043 in the last week. [Image attached]. [ Post]. Mastodon. https://bitcoinhackers.org/@mastodonusercount/109510971085943939

Mastodon Users [@mastodonusercount@bitcoinhackers.org]. (26 October 2022). 5,637,934 accounts +223 in the last hour +3,685 in the last day +18,101 in the last week. [Image attached]. [ Post]. Mastodon. https://bitcoinhackers.org/@mastodonusercount/109236585006366184

mastodon.social. (n.d.). About. Retrieved 14 December 2022. https://mastodon.social/about

Mehnert-Theuerkauf, A. (2022). Trigger warnings and the liberal right to tell people what they do not want to hear. PPmP – Psychotherapy · Psychosomatics · Medical Psychology, 72(07), 281–282. https://doi.org/10.1055/a-1840-7158

Perspective API – How it works. (n.d.). Retrieved 13 December 2022, from https://perspectiveapi.com/how-it-works/

Raman, A., Joglekar, S., Cristofaro, E. D., Sastry, N., & Tyson, G. (2019). Challenges in the Decentralised Web: The Mastodon Case. Proceedings of the Internet Measurement Conference, 217–229. https://doi.org/10.1145/3355369.3355572

Rochko, E. (11 November 2022). Eugen Rochko (@Gargron@mastodon.social). Mastodon. https://mastodon.social/@Gargron/109323056922301691

Roth @systemli.social, A. (@anneroth@systemli.social). (7 December 2022). #Mastodon interim conclusion:There are a few details about Mastodon that make me think it was actually designed as a living room and… [Mastodon post]. Mastodon. https://systemli.social/@anneroth/109473115498414819

Sentiment Analysis. (n.d.). DeepAI. Retrieved 13 December 2022, from deepai.org/machine-learning-model/sentiment-analysis

Schlosser, S. (24 June 2021). Trigger warnings: Don’t use them indiscriminately. Deutschlandfunk Nova. www.deutschlandfunknova.de/beitrag/triggerwarnungen-in-medien-sollten-nicht-inflation%C3%A4r-eingesetzt-werden

Sheehan, E. (23 November 2022). Content Warnings on Mastodon. Darthmall. https://darthmall.net/weblog/2022/mastodon-content-warnings/

Stangl, W. (2022). Werther effect. Online Encyclopaedia of Psychology and Education. Online Encyclopaedia of Psychology and Education. https://lexikon.stangl.eu/10101/werther-effekt

Stefanello, V. (10 November 2022). Assh*le Twitter users are barging into Mastodon and demanding it stop being so polite. The Daily Dot. https://www.dailydot.com/debug/mastadon-content-warnings-twitter/

Thomson, R., Dancy, C., & Pyke, A. (Eds.). (2022). Social, Cultural, and Behavioural Modelling: 15th International Conference, SBP-BRiMS 2022, Pittsburgh, PA, USA, 20–23 September 2022, Proceedings (Vol. 13558). Springer International Publishing. doi.org/10.1007/978-3-031-17114-7

University of Cologne. (2022). Trigger warning. University of Cologne. https://vielfalt.uni-koeln.de/antidiskriminierung/glossar-diskriminierung-rassismuskritik/triggerwarnung

Voit, K. (5 June 2022). Please Don’t Misuse Content Warnings on Mastodon. Public Voit – Karl Voit’s website. https://karl-voit.at/2022/06/05/mastodon-cw-misuse/

Windscheid, L., & Fleiter, P. (n.d.). Crime on Our Doorstep: True Crime from the Neighbourhood (No. 75). https://open.spotify.com/episode/70Fv4EnAlu0ZXloDyPZg0z

Zignani, M. (2019). Mastodon Content Warnings: Inappropriate Content on a Microblogging Platform [Data set]. Harvard Dataverse. doi.org/10.7910/DVN/R1HKVS