In the past decade, English speakers under 35 have developed an expanding vocabulary for low-grade romantic mistreatment. Breadcrumbing — leaving just enough signals of interest to keep someone hopeful without ever committing. Benching — keeping someone in reserve as a backup option. Cushioning — maintaining alternative romantic prospects while in a relationship. Zombieing — reappearing after ghosting as if nothing happened. Orbiting — staying engaged with someone's social media after ending direct contact. Love-bombing — overwhelming a new partner with intensity to accelerate commitment. Roaching — concealing other partners while implying exclusivity. The list grows quarterly. Each word names a specific pattern of behavior that has become common enough to require a label.
The proliferation of vocabulary is itself the story. A culture invents words for the things it sees often. The collective scale is producing these behaviors at scale, and the language is the trailing index. Each new term marks a pattern that has crossed the threshold of being recognizable to a wide audience. This is Law 2 — Think Carefully — applied to behavior: naming a pattern is the first step in being able to refuse it. The lexicon is a small collective intelligence, a distributed effort to make legible what dating apps and the broader medium of modern romance have made common.
What the new vocabulary describes, at root, is the optimization of human attention as a partial good. Each of these behaviors involves giving someone enough attention to retain them as an option without giving enough to actually commit. The behaviors are economically rational under conditions of abundance: when there are always more matches available, holding multiple partial connections costs less than committing to one. The behaviors are romantically corrosive for the same reason: they convert human beings into options in someone else's portfolio. Christian Rudder's data on dating platforms showed that the median active user maintains attention on many more potential partners than they could ever pursue. The new behaviors are the tactics for managing that surplus.
Breadcrumbing is the keystone behavior because it captures the pattern most precisely. A breadcrumber sends a flirtatious text every two weeks. A like on a months-old photo. A "hey stranger" out of nowhere. Each gesture is small enough to deny meaning if challenged and large enough to revive hope. The recipient experiences a recurring small dose of attention that prevents them from moving on. The breadcrumber maintains an option at minimal cost. Marie Bergström's research describes this as "low-investment retention" — the work of keeping someone available without doing the work of actually being with them. The cost is borne by the recipient, who lives in a state of intermittent hope that prevents new connection.
Benching is the structural parallel: someone is kept on a metaphorical bench, available to be brought in if a current option fails. The benched person may not know they are benched. They experience occasional contact, occasional dates, occasional sex, with no apparent escalation. The bencher knows. The bencher is doing portfolio management. Eli Finkel's research on the suffocation model of modern marriage applies here in inverse — modern dating asks people to evaluate so many potential partners that the rational strategy is to never close the evaluation, to keep options warm indefinitely. Benching is the rational outcome.
The new lexicon serves a collective Law 0 function: it gives people the cognitive tools to recognize when they are being mistreated. Without the words, the behaviors are diffuse and easy to rationalize. With the words, the behaviors are nameable, and once nameable, refusable. A person who has been breadcrumbed for six months can, after encountering the term, recognize the pattern and exit it. The vocabulary is a kind of collective immune system, slowly building antibodies to the most common forms of low-grade mistreatment.
The deeper question is whether naming is enough. Sherry Turkle has argued that the digital era has produced abundant language for relational injuries without producing the practices that would prevent them. The lexicon is descriptive, not prescriptive. It tells you what is happening; it does not tell you how to build something different. The work of building something different is the Law 5 work — Revise — that must follow naming. People who learn the words are better off than people who do not. But the people who learn the words are still inside the medium that produces the behaviors. The vocabulary is a partial remedy. The collective practice still has to change.