“I downloaded Grindr,” Logan told me, “and it all changed.” They were talking about the popular location-based hookup app used predominantly by gay men. “I was queer,” Logan went on about their initial experience with Grindr, “and when you’re queer, you have a very different sexual development than straight people. It was so exciting.”
Like Logan, approximately half of American adults between the ages of 18 and 29 use or have used dating and hookup apps. For queer people, though, these apps can present significant challenges: on most, being straight is assumed as default, and few dating and hookup apps make space for queer varieties of attraction. Even apps specifically designed for queer people, such as Grindr, tend to be aimed at a very narrow user base—in the case of Grindr, cisgender gay men. For Logan, who identifies as nonbinary, such apps actually proved to be an obstruction: “It was only when I stopped using dating applications,” they said, “that I figured out that I was nonbinary.”
Part of the problem is that dating applications tend to be designed in ways that force queer users to misrepresent their own experiences. This begins with fundamentals: many of the most popular apps are shot through with elements of binary design, requiring users to select either “woman” or “man” in order to use the application. Yet, for nonbinary users such as Logan, neither option is correct.
The binary assumptions about gender that are currently embedded in these apps help form a web of restrictive assumptions about what kinds of romantic and erotic arrangements are normal, natural, and good. In contrast, many queer subcultures don’t espouse the idea that dating, sex, love, and romance only come as a package deal. Instead, a variety of relational forms, such as “romantic friendships,” are culturally intelligible, such that individuals are well positioned to negotiate what interactions, commitments, and reciprocal relations might be possible.
Designing apps to support a multitude of relational forms, then, would be beneficial to everyone. The point isn’t only to chip out space for queer individuals, but also to share the hard-won wisdom embedded in queerness.
In mainstream Euro-American culture, “woman” and “man” are assumed to exhaust the options for adult gender identity. If an adult individual is not a woman, so the assumption goes, then they must be a man. At work here is a binary gender system. And this system, spread beyond its original boundaries by the workings of colonialism and capital, has been devastatingly harmful around the world. Maintaining this system has required extensive social scaffolding, which (ironically) explains why so many people now take it as a given that the “natural” form of gender is binary.
Yet we know that many cultures, past and present, countenance nonbinary gender identities. “Nonbinary” is a coarse-grained term that can be used to encompass a broad range of identities that, in the words of philosopher Robin Dembroff, exist “in opposition” to the binary gender system “at its conceptual core.” Under the nonbinary umbrella may fall identities such as “agender,” “bigender,” “genderqueer,” and “pangender.” Some transgender people also identify as nonbinary, with identities such as “transmasculine” cutting across the grain of binary gender.
Thinking about the role of masculinity and femininity in everyday life, it is no surprise that gender has an outsize impact on the experiences people have on dating technologies. Women are dozens of times more likely than men to experience harassment on dating apps, an underdiscussed form of abuse. Men and women even appear to use apps differently. A group of researchers recently suggested that dating applications are designed in ways that encourage men to “spread” attention across multiple potential romantic partners and women to “hedge” commitment to any particular potential romantic partner.
Gender also affects the ways in which queer users experience dating applications. GLAAD reports that the profiles of transgender individuals are often maliciously flagged as inappropriate, in some cases causing trans users to be banned from dating and hookup apps “simply for being trans.” Some dating apps are trying to overcome these challenges and make their apps better for queer people. For example, Tinder, one of the biggest online dating platforms, updated its app in 2016 to provide users with more gender identity options, and thus considered itself to have largely solved the problem of transphobia on its service. But trans and nonbinary individuals continue to report unfair treatment on Tinder, and many are still dissatisfied with the design updates it and other apps have offered. I suggest that this is because the core design of these apps remains fundamentally binary, and nonbinary users are simply expected to make due.
Design always pairs technical with ethical considerations. With regard to the technical, design researcher Don Norman, for example, argues that a well-designed door ought to wear its affordances on its sleeve: “here can be pushed,” “here can be pulled,” etc. From a technical perspective, a well-designed dating app would be fast, glitch-free, and intuitive. Yet, in addition to being technically sound, doors ought to be accessible, designed in ways that allow ease of use for a variety of individuals. Likewise, we can ask if a dating application is ethically designed. At minimum, dating applications ought to be transparent about their costs as well as safe, offering tools to protect users against sexual assault, hate speech, and other forms of violence that might result from use. These requirements can be derived from a sparse, baseline conception of ethics. They are perhaps the kind of basic assurances Google’s Code of Conduct has in mind with its injunction, “Don’t be evil.”
Ethical design, however, could involve a more robust vision of what it hopes to make possible—an aspiration to more than simply the avoidance of evil. A core ethical aim in designing technologies, I contend, ought to be nothing less than the promotion of human flourishing. With regard to dating technologies, ethical design might even be given a liberatory gloss. For example, to draw from Amin Ghaziani’s work in urban geography, dating technologies could be designed in ways that promote cohesion among members of communities.
So what then would a truly robust vision of ethical design recommend for the design of dating apps?
Because Tinder is an especially popular dating app—and because it has signaled some of the greatest willingness among such apps to make its product inclusive—I will focus on it in the following discussion. If Tinder really has a good-faith commitment to serving nonbinary people, and queer people more broadly, what should this look like at the level of design?
Essential to Tinder’s design are search parameters that allow users to tailor what sort of profiles might appear in feeds of potential matches. Tinder’s current search parameters enable individuals to pare down their feeds with respect to age, location, and gender identity. These search parameters will have come to feel intuitive to many users.
Yet, various alternatives were possible. Search parameters might have been designed to allow users to filter on the basis of wealth, body mass index, or race. Indeed, in one way or another, many apps do. On the one hand, such search parameters risk reinforcing users’ stereotypes, further entrenching discriminatory patterns of attraction. On the other hand, such filters might help users to pursue erotic solidarity. Recon, a gay kink hookup app, claims that its race filter is necessary so that people of color can find one another: “Without such tools, marginalised people cannot reach out to others who look like them, ask for advice on how to conduct BDSM practises safely, solidify their identities, or in the simplest form, just be seen and feel like they are not alone.” A similar argument might be made in favor of a filter for app users who wish to find partners with whom they share other features or identities, such as being trans or having a large body. Here, I don’t want to rule on the issue, except to emphasize that—whatever the design solution—search parameters are ethically significant.
In its own search tools, Tinder continues to privilege binary gender identities, despite the marketed changes it undertook in the name of queer inclusion. Users on Tinder can only indicate romantic interest in women, men, or women and men—and that’s it. There is no option to indicate interest in nonbinary individuals, let alone the more granular identities under that umbrella. And while users can declare their gender as “nonbinary,” they still must select either “woman” or “man” in response to the prompt: “include me in searches for.” If Tinder’s design were personified, it might say: “You’re nonbinary? Fantastic. Everyone is welcome! Now, should I treat you like a woman or a man?”
Operative in Tinder’s design is a significant misunderstanding of nonbinary identity. Imagine a hypothetical system in which users would be required to select either “woman” or “nonbinary” in order to appear in feeds of potential matches. Even if users had the option nominally to select “man” as a gender identity, the hypothetical design would fail to “treat men like men” in the sense of categorizing men as men in feeds of potential matches. It’s easy to recognize what would be wrong with a dating application in which men don’t have the option to be included in searches as men—but the current design of Tinder makes an identical error with respect to nonbinary identity. Tinder currently offers token recognition to nonbinary people without any real acceptance, representation, or accommodation.
That this kind of mistake was even made speaks in part to the severe marginalization of nonbinary people in the culture at large. For example, Sasha Costanza-Chock, who works as a community-based design theorist, describes how binary design affects their experience of moving through airport security:
My heartbeat speeds up slightly as I near the end of the line, because I know that I’m almost certainly about to experience an embarrassing, uncomfortable, and perhaps humiliating search by a Transportation Security Administration (TSA) officer, after my body is flagged as anomalous by the millimeter wave scanner. I know that this is almost certainly about to happen because of the particular sociotechnical configuration of gender normativity . . . that has been built into the scanner.
Individual TSA agents might act differently in response to “anomalous” reports from the scanner, but what’s needed is a technology that does not treat gender deviance as a security threat. The underlying issue here is binary design. Technologies ought not to represent “woman” and “man” as the only intelligible or otherwise legitimate gender identities.
For its part, Tinder claims that its reliance on binary search parameters is about foregrounding user experience. Former CEO of Tinder Sean Rad provides the following argument in favor of the current design: “We had to keep it simple; if we didn’t keep it simple, no one would use these features.” But simple for which users? The current search parameters are tailor-made for women and men who are exclusively interested in dating other binary individuals, while misrepresenting everyone else.
In the abstract, the design solution seems straightforward. If men have the option to be included in searches for men, then nonbinary individuals ought also to have the option to be included in searches for nonbinary individuals. And if users (of any gender identity) have the option to search for men, then users (of any gender identity) ought also to have the option to search for nonbinary individuals.
In practice, however, the solution isn’t so clear-cut. Here are a few designs that would, in various ways, aim at equal treatment of queer and straight individuals. Each option comes with distinct costs and benefits, and I offer each design as a thought experiment, not as a model ready for implementation.
Proposal 1: No gender categories
A design that rejects the drive to name and classify might be celebrated by canonical figures in the field of queer theory such as Michel Foucault and Eve Kosofsky Sedgwick. For example, Maggie Nelson provides the following reflection on the latter figure: “Sedgwick wanted to make way for ‘queer’ to hold all kinds of resistances and fracturings and mismatches.” Categories, which essentially unify and stabilize, stand in contrast to the aforementioned vision of queerness.
Along these lines, dating applications might want entirely to forgo user-selected gender identities and preferences. A group of my students in a seminar on the philosophy of sex, sexuality, and gender, for example, proposed a design in which users are randomly displayed profiles, with an algorithm tailoring feeds of potential matches over time. The algorithm, presumably, would operate upon data related to indicated interest in other users, conversation length among mutually interested users, and rates of eventual offline interaction.
In case that sounds dystopian, it’s interesting to note that an algorithm might be designed, as suggested by a group of researchers at Cornell, to reduce bias within feeds of potential matches. Plus, as Rebecca Solnit has compellingly explored, we often do not know what we want until we find it. As information technologies continue to develop, the search-and-swipe format might eventually become obsolete.
At least when apps are used in binary cultural settings, however, a category-free design might nonetheless indirectly reproduce binary gender assumptions. As philosopher Gabbrielle Johnson has explained, algorithms need not directly operate on data about group membership in order to function as if they did—given that other data points can serve as proxies for group membership. This general point is already familiar from discriminatory redlining practices in which zip codes serve as proxies for race. If an algorithm were only to sort “masculine-presenting” nonbinary individuals into a group that predominately included men, then the design wouldn’t amount to much of an improvement over the current system on Tinder. Safeguards against this issue of indirect binary categorization would need to be put into place.
Proposal 2: Single-level design
If, instead, we work from the idea that dating apps ought to include gender-based search parameters, what sort of design would accommodate the lived experience of queer and nonbinary people? What categories ought to be available? And how ought gender identities be organized and displayed?
One option is a single-level approach in which users select from an inclusive array of gender identities. There are, indeed, a few small (compared to Tinder) queer-specific dating apps that use this very design strategy. For example, consider Taimi, which markets itself as “the world’s largest LGBTQ+ social and dating app.” Taimi enables users to identify as “male,” “female,” “trans male,” “trans female,” “intersex,” or “nonbinary,” and the same categories are available to filter feeds of potential matches. Importantly, Taimi displays gender identities in a symmetrical fashion, as part of a list in which each identity is given equal weight. Users don’t need to select an option such as “other” in order to then further select “trans” or “nonbinary.” Similarly, Feeld, advertised as “a space where you can explore your identity,” allows users to select from over a dozen gender categories, which are also available in search functions. Notably, Feeld includes an option, “trans human,” which is nearby to a possibility in which users might entirely opt out of selecting a gender.
But any single-level approach confronts a tension between the technical and ethical aspects of design. A single-level design, unless it includes an expansive list of categories, will risk misrepresenting the gender experience of at least some individuals. Consider gender identities such as “transmasculine” and “transfeminine,” which resist clear-cut categorization as “man,” “woman,” or “nonbinary.” For example, in photographer Soraya Zaman’s American Boys, a series of portraits of transmasculine individuals, Caden, photographed two years after starting to take testosterone, explains: “I’m very masculine but I’m also very feminine and I’m on the spectrum for sure. Being on testosterone has made me more comfortable with myself and I’ve been exploring both my gender and sexuality.” A dating application ought not to require (while nonetheless allowing) individuals such as Caden to identify as male or nonbinary.
The app OkCupid currently provides twenty-two gender options, including “transmasculine.” Moreover, OkCupid’s design uses check boxes that enable individuals to identify with multiple genders. In contrast, dating applications including Tinder, Feeld, and even Taimi use radio buttons that require users to select a single gender. Radio buttons might be apt for dealing with information such as height, given that it is not possible to be five feet tall and six feet tall at the same time. But gender is not like that: many queer individuals have multiple gender identities.
Nonetheless, Taimi, Feeld, and especially OkCupid all offer radical improvements over the binary design of applications such as Tinder, even if none provide an exhaustive array of options for gender identity.
Proposal 3: Multilevel design
Recall that gender categories such as “bigender” are fine-grained, constituting specific identities. In contrast, gender categories such as “woman,” “man,” and “nonbinary” are coarse-grained, potentially encompassing various fine-grained identities. Dating apps that use categories ought to be designed in ways that enable individuals to identify with gender categories at each level of granularity.
In contrast to a single-level approach, on a multilevel approach, users could select gender identities via top-down or bottom-up mechanisms. For example, in a top-down fashion, a user might select a coarse-grained identity such as “nonbinary.” Next, the user would have the option (without being required) to identify with fine-grained categories such as “agender,” “bigender,” “genderqueer,” and “pangender.” A top-down selection mechanism would accurately represent individuals who don’t identify with any fine-grained gender categories.
Or, in a bottom-up fashion, a user might select a fine-grained identity such as “transmasculine.” Next, the user would have the option (again, without being required) to select among coarse-grained identities such as “nonbinary” and “man.” A bottom-up selection mechanism would accurately represent individuals who don’t identify with any coarse-grained gender categories.
Multilevel design could be either integrated or modular. On an integrated approach, if a user were to select a fine-grained identity such as “bigender,” the dating application would automatically categorize their profile as nonbinary. An integrated design would need to provide users with something like the following option: “Do not include me in searches for women, men, or nonbinary individuals.” In contrast, a modular approach wouldn’t automatically extrapolate about the gender identity of any particular user across levels of granularity.
Any version of the multilevel approach would ideally prompt something like translation. Representing gender identities across levels of granularity could facilitate interpersonal and cross-cultural intelligibility, all while making space to honor the many ways in which individuals find meaning in specific terms.
A multilevel design would make salient a series of questions about gender and sexuality: What does it mean to be a man? To be straight? Am I attracted to bodies? Stylizations? Social roles? What influence, if any, do I have over my attractions?
Admittedly, cisgender straight individuals might not want to confront these questions. Applications such as Tinder were undoubtedly developed with straight users in mind, and their designs insulate users from such considerations. With exasperation, it might be added that the same point applies to nearly everything in mainstream culture.
As part of other research, I’ve advanced a sort of ambassadorial idea: we can make the world (more) just by making it (more) queer. A queer world is a world in which everyone is quite a lot happier. Straight people currently navigate so many intensely rigid binary norms about who to be and how to relate to each other. Fortunately, queer people have created a variety of alternative models—ways of life so liberating that generations of activists have endlessly pursued their realization, even at the cost of stigma, erasure, and violence.
We have a right to demand equal treatment and accurate representation in the context of dating applications. But these changes are universally beneficial. Instead of compressing love, sex, and care into prepackaged forms, ethically designed technologies could enable queer and straight individuals alike to freely express identity and desire. Ethical design has transformative potential.