In this extract from his forthcoming book, Alfie Bown speculates as to what a leftist dating simulator might look like. Dream Lovers: The Gamification of Relationships was out this month with Pluto Press. This week, Alfie was on Diet Soap Podcast with Doug Lain discussing the book, an analysis of the ways in which sex tech – from dating sites to smart condoms, sexbots, and VR pornography – is transforming desire today.
Heard of Trump.dating or tried Red Yenta? The Atlasophere? Christian Mingle? Probably not, but Hinge, Bumble, Tinder, Grindr, and OkCupid are probably more familiar. Outside the ‘Match Group’ of sites owned by IAC, there are dating sites which attempt to be for those on the political right and even for the liberal dater, but are there any dating sites that are really, seriously, for the Left? Will there even be dating apps after the revolution? What would one look like anyway?
In her polemical 1970 feminist text The Dialectic of Sex, Shulamith Firestone presents the idea that when it comes to love women cannot afford to be spontaneous. That, she argues, if often the privilege of those in structural positions of power, in the context of her argument: the position of men. This is important to remember when it comes to the discourses that surround dating applications. Those who criticise the world of online dating usually argue that offline or ‘traditional’ dating offers a more organic, natural and spontaneous way of meeting a partner. Firestone’s point raises the question of whether that would be in any way desirable, especially in a world in which economic conditions and precarity make every second of spare time precious and especially given those traditional structures of dating have always included inequalities.
On top of this, it’s not at all certain that offline or pre-internet relationships are any more organic and spontaneous than the algorithm-driven ones we experience today. In The Purchase of Intimacy Viviana Zelizer discusses the kinds of legal and social contracts which have long since functioned to organise and sort people into couples:
Outside the legal arena, in ordinary, everyday practice, people engage in a similar sorting of couples. They do not employ precisely the same distinctions as lawyers or invoke exactly the same moral evaluations of different kinds of relations. But they sort across the whole range of relations that involve the possibility of intimacy, from lawyer-client or doctor-patient to friends, neighbours, workmates, and kin.
Social life functions to sort individuals into couples along economic, social and political lines. As we saw in the first chapter, dating applications inherit some of these social biases and embed them into algorithms whose datasets then go on to set the terms for friendships and relationship of the future. In light of the fact that it is perhaps both inevitable (Zelizer) and desirable (Firestone) to organise the political processes by which coupling takes place, we might playfully suggest the blueprint for an online dating app which might organize couples in more politically and economically desirable ways than the existing set of applications out there.
To make such a program, we’d need to think both about its algorithms and its interface. When it comes to the algorithm – the basic way it matches users with each other – we could start by describing the existing grouping structures used by typical matchingalgorithms and by thinking about how we might want to structure these processes differently. Since both Roland Barthes and Sigmund Freud have been used regularly in this book, we could playfully use the distinction that Barthes makes between “metaphor” and “metonymy”, which relate to Freud’s earlier ideas of “condensation” and “displacement”, to differentiate the way dating apps and algorithms are from the way in which we might want them to be. As things stand now, we can say that dating apps work by a process of “metaphor”.
Metaphor and metonymy have often been considered two fundamentally different ways of humans structuring their discourses and ways of thinking. With metaphor, there is an impression of similarity (hence the longstanding association of metaphor with a simile in the classroom) between the two items or objects in the process which are directly compared, as in Romeo’s ‘Juliet is the sun!’ or in the world’s most famous metaphor, ‘shall I compare thee to a summer’s day?’. In effect, this logic of similarity is the basis for dating sites and their algorithms, as well as – though perhaps to a lesser extent – for friend suggestions on social media networks. The logic of such algorithms is to organise people into large sets and connect them with individuals whose data is similar. Trump.dating is an exemplary case.
Even in the interface of dating platforms, the logic is fairly clear: we select categories from a drop-down menu or answer extensive multiple-choice questionnaires which puts us into a pool with other uses who selected these answers. On Hinge you can identify your politics by picking between liberal, moderate, conservative (notably omitting any leftist position), while on Bumble you can identify as liberal, right, moderate or apolitical. Other questions are less directly political, such as OKCupid’s “Do you like the taste of beer?’ or ‘Would you like to be the Supreme Ruler of Earth?’. The OKCupid official blog explains the process, saying ‘if you answer “no” to the question “Should the government defund Planned Parenthood?” wouldn’t you want your potential date to answer the same way?’. The algorithm works by finding matches with the closest similarity to you, seeing the ideal match as a kind of reflection of you. In other words, there is a metaphorical logic of direct comparison between the two objects/lovers.
Opposed to this logic of connecting two things, Barthes discusses the idea of metonymy. Metonymy is often used casually to mean the substitution of a whole with a part – ‘the part taken for the whole’ as when the word ‘suit’ is used to describe a businessman. Rather than being based in similarity, the logic here is that there is an association between the two things. Another word for metonymy is contiguity, meaning to border upon something or to touch something. Lacan adds to this discussion in his famous essay ‘The Instance of the Letter in the Unconscious’ where he re-connects metaphor and metonymy with psychoanalysis. Here he suggests that metonymy has a more important connection to language, saying that ‘the connection between ship and sail is nowhere other than in the signifier’ and that therefore metonymy ‘is based on the word-to-word nature of this connection.’ In other words, metonymy is more like the children’s word association game where players must shout out the first word that comes to mind when another word is spoken: it is not so much about a similarity between the two words but about an association that exists in the unconscious that is made visible by the effects of language.
This could form a blueprint for designing an algorithm for a dating site that was able to cut through the kinds of social organization that is common in existing sites. It might create a way of connecting people that uses data to make helpful suggestions while supporting a diverse range of class and cultural connections that could reach outside of the filter bubbles produced by the splinternet of cyberbalkanization. If suggestions were based on metonymic connections between language used rather than similarity between users, all the datainput into the site by the user would be taken into account and deployed, but the process would not preclude connections with those dissimilar to themselves.
This might lead to occasionally unhelpful suggestions. For instance, if the word ‘left-wing’ it would be proximate metonymically to ‘right-wing’ even though those two users might be unlikely to match, making the site rather the opposite of Trump.dating. However, since the algorithm would use hundreds of words rather than one, such a connection would be unlikely. Rather, users would be connected to those speaking in related terms to themselves. Wordmaps have been trending on Facebook and Instragam for the last several years as a means of reflecting on user’s discourse, and such maps could form the basis of a dataset that would not seek to connect the user with others using the same words but with related words in a huge big data map – a kind of word association game with millions of players – and aggregated into a model for making connections between people. In such a model, you would not be connected with those like you but those whose discourse connects to or intersects with yours.
In terms of how the interface might work, lets imagine a dating site structured like Reddit or like Wikipedia – a platform which works very much like the word association game (just think of the hours spent clicking thorough the rabbit holes of different pages and links). Such an experience would be almost the complete opposite to a swiping platform like Tinder or Hinge. Importantly, both are game-like, but they are completely different kinds of game. In one, you are a kind of digital detective, an amateur sleuth going through these links and holes, searching for connections and a place to rest your attention for a time, actively pursuing links that might be more likely to lead to that outcome. In the other, you are entirely passive and powerless as to which page will appear next, beholden only to the logic of the application.
Each page on the site could be hosted by a single user, equivalent to their profile page on a dating site, and every word used on that page could be hyperlinked to other pages that have used that word or words metonymically connected to it, so that the user can jump from one profile to another by identifying a word or cluster of words of interest. For the most part, this would keep users in loops and bubbles of users who they share interests and connections with, but it would not preclude the user from actively moving outside of that bubble for any distance if they chose to pursue a particular line of interest of enquiry. Like with an hour spent on Wikipedia, users will find themselves far from there they usually are, crossing the boundaries of the splinternet and its organization capitalist logic.
For the most part this is a playful rather than a serious suggestion, but it hopefully at least points the need for solutions to the problems of connecting in a data-driven society such as ours. It’s clear that we need data and, as Firestone wrote, cannot always afford to be spontaneous. However, we also need ways to ensure we are not reduced to manageable bubbles of individuals partitioned off from each other by algorithms and curation tools which work by organizing us according to class, race and gender demographics. If we are to foster greater solidarity with each other across cultural, economic and social borders, we need first to be able to see each other, and today that means appearing in each other’s digital worlds.