Computer-Mediated Consent to Sex: The Context of Tinder
This blog post summarizes a paper being presented at the 24th-annual ACM Conference on Computer-Supported Cooperative work and Social Computing (CSCW 2021), and published within their proceedings. Written by Douglas Zytko, Nicholas Furlo, Bailey Carlin, and Matthew Archer, the paper examines consent navigations in digital context, and discovered two distinct processes: consent signaling and affirmative consent. The paper has won both a Best Paper and an Impact Recognition award from CSCW, and may be read in its entirety here.
CONTENT WARNING: This paper directly discusses (and quotes participants) on issues of non-consensual physical/sexual activity/sexual violence.
How is consent to sex (or permission to a sexual activity) exchanged between sexual partners? You probably wouldn’t say “with an app,” but we found that the ways that people give and (perceive to) receive consent are mediated by prior interactions and content on dating apps — and these computer-mediated consent practices can lead to sexual violence. In this blog post we summarize our findings about how consent to sex is computer-mediated, and how this knowledge can transform dating apps from a sexual violence risk factor into a scalable solution to sexual violence all across the world.
Consent exchange is often conceptualized as something that happens in-person, such as by asking for, and sharing, verbal agreement right before sex. But just as technology has given us new opportunities to discover sexual partners, it has also given us new opportunities to (mis-)communicate consent to sexual activity.
Our primary research question was: how is consent to sex computer-mediated?
This is a pressing question because of the severity of sexual violence, or sexual activity without consent, particularly against women and the LGBTQIA+ community. Sexual violence has become a topic of interest in Human-Computer Interaction (HCI) both in regards to understanding how computer-mediated communication perpetuates the problem, and how new technologies could be designed to stop it.
With an understanding of how sexual consent practices are — perhaps inadvertently — mediated by computers, we can design new technologies to intentionally mediate consent practices in ways that mitigate risk of sexual violence.
To explore our research question, we conducted interviews with users of dating apps, particularly Tinder. Nearly 50% of people under the age of 30 have used dating apps. Dating app-use has also repeatedly been correlated with sexual violence. Our participants represented a variety of gender and sexual identities, as well as cross-sections of college student communities (such as Greek Life).
We discovered two distinct computer-mediated consent processes, which we call consent signaling and affirmative consent.
With consent signaling, Tinder’s interface was used to infer and imply consent to sexual activity without any verbal or overt confirmation before the activity was engaged in. These users understood Tinder to be an app intended to support discovery of sexual partners and rapid progression to sexual encounters. Profiles were thus perceived as tools for identifying nearby people interested in sex, rather than tools for evaluating if someone was on the dating app for sex. A “match” in Tinder’s interface (when two users “swipe right” on each other’s profile, enabling them to exchange messages) was considered a signal of mutual sexual interest and consent to sexually explicit messaging. Agreement to meet face-to-face through the messaging interface was then interpreted as agreement to have sex. Experiences of consent signaling were predominantly from users identifying as heterosexual men and women, and some LGBTQIA+ users when reporting on their early sexual experiences through Tinder with heterosexual partners.
With affirmative consent, Tinder’s interface was used to prompt overt discourse around consent before face-to-face meetings. The term affirmative consent comes from public health literature and legislation that rejects the lack of resistance to sex as a form of consent, and necessitates that partners give unambiguous agreement to sex (“yes means yes”). Tinder’s interface was conceptualized by some participants as affirmative consent software, enabling them to foster patterns of affirmative consent and identify resistance to affirmative consent practices in a reduced-risk computer-mediated environment. Consent to specific sexual acts would be overtly exchanged before ever meeting a sexual partner face-to-face, and then verbally reconfirmed during the face-to-face meeting. Experiences with affirmative consent were entirely from users identifying as LGBTQIA+ and were typically adopted in response to sexual violence or general harm that such users experienced in their prior dating app-use.
Both of the discovered computer-mediated consent processes pose risk of sexual violence.
… participants described letting a sexual experience happen that they otherwise did not want for reasons including a perceived obligation to have sex and fear of physical retaliation.
In no experience of consent signaling was consent verbally confirmed prior to a physical sexual activity occurring. This poses a risk of nonconsensual sexual activity because of the possibility that signals of consent could be misinterpreted. The initiators of physical sexual contact through consent signaling — all cisgender men — also conveyed problematic conceptualizations around consent. For one, they believed that their partners did not want to overtly discuss consent. In lieu of verbal discussion of consent, cisgender men believed that they could “sense” when their partner was ready to have sex. This was typically signaled through a lack of physical resistance to a sexual advance. Other participants described letting a sexual experience happen that they otherwise did not want for reasons including a perceived obligation to have sex and fear of physical retaliation.
Despite computer-mediated affirmative consent being adopted by users specifically to mitigate nonconsensual sex, participants exhibited reasons why this process can still expose users to sexual violence. For one, use of the dating app for scaffolding affirmative consent stops once users transition to face-to-face encounters. Once-overt dialogue about sexual boundaries and consent over asynchronous messaging can make way to ambiguous face-to-face situations where the need to (re-)exchange consent becomes “blurry,” leaving victims to hesitate in moments of nonconsensual contact. Participants reported additional struggles with trying to implement computer-mediated affirmative consent through the dating app. While disclosing sexual desire in their profiles sometimes helped trigger conversations about sexual boundaries, other times it was misinterpreted as an invitation for sexual objectification.
What if dating apps were intentionally designed to mediate sexual consent exchange in ways that mitigate sexual violence?
The silver lining: while the discovered consent practices leave users susceptible to sexual violence, this computer-mediation of consent is not necessarily intended by designers. What if dating apps were intentionally designed to mediate sexual consent exchange in ways that mitigate sexual violence? Future design in pursuit of this question could transform dating apps into incredibly powerful sexual violence prevention tools.
Zytko, D, Furlo, N., Carlin, B., & Archer, M.W. (2021). Computer-mediated consent to sex: The context of Tinder. Proceedings of the ACM on Human-Computer Interaction, 5(1), 1–26. https://doi.org/10.1145/3449288
