Several years ago I was walking a team through the results of a study testing prototypes with twelve PC game players. Responses had varied, but what I was excited to share was that across these gamers, I had identified a range of distinct motivations for play. I never got to that part of the presentation. Within five minutes, an argument broke out between the organization lead and his second in command about the underlying technology and what to do about a program I had never heard of. As I tried to get the meeting back on track, the organization lead asked me, “Why does this matter?” “How is this useful?”

I learned a few things that day. For one, a “key takeaways” slide at the beginning helps a lot. More importantly, for product and solution teams, understanding is a means to an end, not an end in itself, and “meaning” is less meaningful than implications, which are in turn valuable only insofar as they point toward actions.

It is easy to scoff at such attitudes as willful ignorance. Yet I have come to see this seeming lack of regard as symptomatic of my own gap in understanding what mattered most to my colleagues and why. The people I work with are faced every day with making concrete and consequential decisions in ambiguous circumstances. Competition is heating up, what should we do? We have a technology that will cost millions to improve by x percent—is that investment warranted? We need to finalize a feature list now, though the product won’t be out for years; which functions are critical? At risk is not just profit but jobs and livelihoods. Given this, it is unsurprising that stakeholders have little patience for insights pointing out complexities, that they yearn for simplicity and certainty. Anthropology, often focused on teasing apart enmeshed and interrelated structures of power and meaning, can seem a mismatched partner indeed.

And yet, while my stakeholders don’t have much bandwidth to appreciate what Clifford Geertz (1973) said about the stakes of deep play, or how Richard Schechner (2002) complicated notions of play across boundaries of awareness and consent, I would argue that anthropological theory has critical contributions to make. More, it is the embedded anthropological theory and the habits of critical analysis that make my work effective.

In 2018, research I led identified “toxicity”—harassment, abuse, and disruptive behavior—as not only concerning for video game players, but a limiting factor on the PC gaming market. This set off a series of activities to gauge that threat and what we should do about it. The team decided to focus on detecting toxic language in voice chat. Initially, this was envisioned as a “mute” button that would empower players to control their exposure to others.

One of the things that makes harassment in gaming difficult to pin down is the way it is situated within play. People who abuse others accuse those they have harassed as not understanding that “it’s a game.” Racist, sexist, nationalist, and otherwise hateful behavior is framed as humor where those who react are cast as overly sensitive and thus “not real gamers.” Clearly there is disingenuousness here but, importantly, toxicity in gaming often relies on the relationship between the “twitch” and the “wink,” of reflex and meaning, and on the knife edge framing of what is and isn’t play—and for whom.

Players themselves are often ambivalent. On the one hand, abuse is real, and a substantial number of people we interviewed had suffered—sometimes severely. On the other hand, nearly all players told us told us it really depended not only on what was said, but who was speaking. Name calling between trusted friends, for example, can be fun; abuse from unknowns can be deeply upsettling. Yet even these lines can be fuzzy. Friends cross lines; witnessing cruelty without context can affect the audience, even where the participants are having fun.

We used these observations to drive concrete conversations within the project team on the importance of enabling users to have more control over what is and is not filtered and when. We also worked to lay out the broader complex of policies, metrics, terms of service, and game features within which our solution operates so that we could be clear as to how it fits today and what it might offer down the road.

By teasing out ways that toxicity and harassment unfold, and the interpretive frameworks gamers use to understand their own experiences, the team was able to craft a more nuanced approach. Today’s solution allows users more choice as to kinds of language they want filtered, and enables reviewing what they didn’t hear, after the fact, with options for reporting it. Ultimately, we would like to enable “safe” people such that gamers can make decisions based not only on what is said but also who is speaking.

The project is now in testing, but what I hope emerges is that while Geertz (1973), Gregory Bateson (1972), and Schechner (2002) never appear in work I shared with my team, insights were drawn not merely from method but from theory. Theory enabled us to see patterns running through the stories we heard. These, in turn, facilitated recommendations and conversations that dug deeper into the social context and implications of “play” for players. What made this project different from the one so long ago was not the presence or lack of theory but the focus on what our findings meant for the decisions the team needed to make and why.

Social and civic life is increasingly digital, unfolding across multiple fragmented networks of private and public domains. There are myriad problems this engenders. Anthropological theory has the potential for great impact if, as that executive once asked, we can make explicit why it matters and how it is useful, to risk our critical stance into concrete actions.