The Trouble with Trigger Warnings

My friends and readers are familiar with my perennial argument: RPG texts are technical documents. They’re instruction manuals that guide us through the process of playing a game.

Standards dictate that instruction manuals containing potentially dangerous procedures must include relevant, specific safety warnings. Technical writers should present these warnings up front as well as at relevant points in the text to emphasize risks and enable users to safeguard themselves against harm.

For exactly that reason, content warnings have become commonplace in RPG materials and marketing copy. Designers want to help players avoid undesirable and psychologically traumatic effects during an experience intended for fun and entertainment.

But do warnings actually fulfill this purpose?

In September 2021, Momatoes began secondary research on trigger warnings’ effectiveness, and she shared a digest of her findings with me. I was able to obtain (with some greatly appreciated assistance) the more recent papers she uncovered.

Although the studies have limitations (discussed below), the results are fairly unambiguous: researchers have consistently found warnings don’t perform as intended. Instead, evidence suggests warnings actually trigger and intensify stress responses instead of helping prevent or mitigate them. They may also encourage countertherapeutic avoidance and reinforce centrality of trauma in personal identity, both of which are harmful to long-term wellbeing.

A (Very) Brief History of Trigger Warnings

Trigger warnings originated in online communities of trauma survivors. Members prefaced their posts with warnings about content that could induce post-traumatic stress responses. They wanted to avoid blindsiding survivors and other readers with this content, allowing them to prepare themselves for engagement or to choose to avoid engagement (and therefore distress) entirely.

Trigger warnings have since spread to other venues like classrooms and social media, and they’ve become a mainstay of the TTRPG community. In all cases, they’re used with nothing but the best intentions.

However, we must recognize that trigger warnings’ origin is informal. They were not developed and deployed by professionals in clinical settings. Their efficacy has been entirely anecdotal—until recently.

The Primary Research

I’ve reviewed three relatively recent papers that evaluate content warnings’ effectiveness and outcomes. The findings are summarized below.

Trigger Warning Efficacy: The Impact of Warnings on Affect, Attitudes, and Learning

2019. Guy A. Boysen, Raina A. Isaacs, Lori Tretter, Syndie Markowski; McKendree University

This report focuses on trigger warnings in educational settings. Teachers deploy trigger warnings under the assumption that students inevitably struggle to learn while experiencing psychological distress. But the researchers found no evidence that trigger warnings influence emotional regulation.

Participants who received trigger warnings before engaging sensitive content tended to believe more strongly that content warnings perform an important role—but only for other people. They generally did not perceive a personal need for trigger warnings.

The researchers advise teachers to be aware of the severity of sensitive course material and students with relevant personal histories, but they conclude that trigger warnings are unlikely to impact most students’ levels of distress or ability to learn. They recommend selective application rather than blanket usage.

In sum: trigger warnings tend to reinforce the belief that trigger warnings are necessary (at least in some circumstances), but they aren’t found to influence emotional regulation in response to sensitive content.

Trigger Warnings and Resilience in College Students: A Preregistered Replication and Extension

2020. Benjamin W. Bellet and Payton J. Jones, Harvard University; Cynthia A. Meyersburg, Foundation for Individual Rights in Education; Miranda M. Brenneman, Coastal Carolina University; Kaitlin E. Morehead, Coastal Carolina University and Foundation for Individual Rights in Education; Richard J. McNally, Harvard University

This study examined a student population without a history of trauma. It found substantial evidence that trigger warnings cause small increases in immediate anxiety and fail to reduce anxiety responses to distressing content.

The researchers provide a sprawling literature review of related findings:

  • Trauma survivors who associate acute symptoms as signals of enduring disability are at greater risk of developing PTSD. The belief in trigger warnings’ absolute necessity undermines the general perception of survivors’ ability to cope with stressors, whereas most survivors experience dissipating symptoms over time.
  • Gradual exposure to trauma cues allow survivors to habituate and regain control of functioning. Avoidance is associated with more severe PTSD symptoms.
  • Trigger warnings increase perceptions of vulnerability to sustained impairment and don’t affect immediate anxiety responses.
  • Studies conducted using visual media found trigger warnings increased anxiety. In contrast to situations where those warnings weren’t presented, trigger warnings induced negative expectations but didn’t appreciably influence negative effects.
  • Trigger warnings were shown to slightly reduce subsequent negative affect when engaging texts. Participants who believed trigger warnings serve a protective function were more likely to experience enhanced negative affect and more likely to avoid the material altogether.
  • Trigger warnings trivially increase negative affect in traumatized individuals.
  • Trigger warnings increase physiological response in trauma survivors according to the severity of their PTSD. As a result, trigger warnings may temporarily worsen the wellbeing of the people they’re intended to help.
  • Trigger warnings may encourage countertherapeutic avoidance that reinforce trauma’s centrality to the survivor’s identity. Studies found positive associations between: 1) the quantity of trigger warnings presented and the severity of PTSD avoidance response; 2) trigger warnings and perception of trauma’s centrality to the survivor’s identity.

Helping or Harming? The Effect of Trigger Warnings on Individuals with Trauma Histories

2020. Payton J. Jones, Benjamin W. Bellet, Richard J. McNally; Harvard University

This study investigated whether trigger warnings help survivors cope with (rather than avoid) triggering content. It found that warnings did not reduce anxiety for people diagnosed with PTSD, and the warnings had a trivially small effect on overall anxiety response. When effects emerged, they tended to be increased anxiety rather than the desired decrease. Individuals with more severe PTSD experienced greater anxiety in response to trigger warnings.

The researchers found substantial evidence that trigger warnings reinforce trauma as a central component to survivors’ life narratives. They also found evidence that trigger warnings reinforce (whether explicitly or implicitly) the notion that trauma survivors are uniquely vulnerable. The authors reiterate that increasing trauma centrality is directly countertherapeutic.

Relevant cited findings include:

  • Avoidance reduces anxiety in the short term but maintains PTSD and potentially worsens it in the long term.
  • Progressive exposure to trauma cues over an extended period is beneficial to wellbeing, particularly in a controlled treatment setting.

In sum: trigger warnings reinforce survivors’ beliefs that their traumas are central to their identities, and warnings are therefore countertherapeutic. Evidence favors the hypothesis that trigger warnings have no effect and are not helpful for trauma survivors. Whether or not trauma warnings are actually harmful remains unclear.


The studies summarized here are not exhaustive, and each has limitations discussed in the respective papers.

The most significant limitation for our purposes is that none directly treat roleplaying games or the TTRPG community (which are themselves significantly understudied). RPGs are more interactive than the media and settings in these studies. This does not invalidate the studies’ findings, but it does certainly qualify them.

How Can We Do Better?

That’s the $10 million question.

Roleplaying is a common practice in therapy, and videogames have become a valuable clinical tool, so RPGs certainly have therapeutic potential. However, this is the domain of qualified professionals in a clinical setting. In any other situation, games present the possibility of doing just as much harm as good, and a layperson is unlikely to be able to manage the situation productively.

One suggestion (put forward in “Trigger Warning Efficacy”) is to provide an honest summary of content instead of presenting trigger warnings as such. But like warnings themselves, this runs up against the problem of individuality; it’s impossible to know what may trigger any given person.

In RPGs, safety tools are a common complement to trigger warnings. These tools help every player maintain agency and navigate potentially distressing or traumatic content in games. However, safety tools are currently in the same situation trigger warnings recently occupied: they are tools created by laypeople, not clinical professionals, and their efficacy is supported only anecdotally, not empirically.

Remember: the research shows that people presented with trigger warnings believed they effectively fulfill a valuable purpose, whereas additional research shows that these warnings increase anxiety and may negatively impact long-term wellbeing. Safety tools may instill the same false perception in their users. No research has been conducted, so no judgment is possible.

In the final analysis, the research discussed here does not give concrete, actionable suggestions for how designers can help players navigate difficult content. That’s a complex problem, and complex problems rarely have simple solutions. The biggest takeaway is that using trigger warnings because we assume they’re beneficial may do the community more harm than good. Even if we don’t know how to resolve the problem in the short term, we can at least be aware of the pitfalls and look for new and better solutions.

Liber Ludorum is entirely reader-funded. Please consider lending your support.

The hazard stripes used above are from MyFreeTextures and used under the terms of its license.

Recent posts

2 thoughts on “The Trouble with Trigger Warnings

  1. Really cool to see empirical research! This is so lacking in the TTRPG community, thank you for sharing this! Unfortunately I no longer have my academic google account (or at least I haven’t checked it in a long time 0.o…) so I only was able to access that third paper, but skimming over it, it looked reasonable to me. I appreciate that they used preregistration and were clear about exploratory vs. confirmatory analyses.

    I don’t know enough about the specific academic area the authors are coming from to know what their standard practices are- I haven’t read too many papers like this that use Bayes Factor in a case like this, and I’m not sure why they used that over normal hypothesis testing approaches when nothing they did seemed otherwise all that radical, but it was well explained and nothing about it seemed obviously problematic to me, although again I only gave it a quick look over right now. I am sometimes a little skeptical of how Bayesian stats get used in academia, but I’m too rusty on this stuff to have strong opinions on it in this case lol. I used to study the cognitive neuroscience of language and people did use Bayesian stats, but usually it was due to some stronger theoretical constraint or specific application in mind than what I would think would be necessary here.

    Critiques of their methods or experimental design aside, principally, this is why it’s so important to have empirical research. Because no matter how well meaning, sometimes even the most reasonable theories can be wrong and their applications have no effect or a counterproductive effect, and that can be a lot of effort or cost spent for nothing, or can give people a false sense of security.

    This is awesome, I’d love to see more posts like this!


    1. Thank you! I’m glad this was such a beneficial post. The technical and methodological nuances are outside my wheelhouse, so I’m just going to have to trust their peer reviewers. Luckily, the findings are pretty straightforward and accessible to a non-specialist like myself. The biggest problem is, like you said, that the information tends to be locked behind an academic paywall, so the creators who would benefit from it may never even know it exists.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: