Ray Oldenburg coined "third place" in 1989 to name the category of social space that is neither the first place (home) nor the second place (work). The cafe, the pub, the barbershop, the bookstore, the church basement, the bowling alley, the diner, the post office line in a small town. Oldenburg's argument was civic, not romantic — he was worried about American democracy losing its informal commons. But his framework turns out to be the single best lens for understanding where couples used to meet, and why they don't meet there anymore.
A third place has roughly eight features: it is neutral ground, it is a leveler across class, conversation is the main activity, it is accessible and accommodating, it has regulars, it keeps a low profile, the mood is playful, and it functions as a home away from home. Notice that every one of these features is also a condition for accidental romantic encounter. Neutral ground means strangers can approach each other without trespass. The leveling means a doctor can flirt with a carpenter without anyone keeping score. Conversation as main activity means there is permission to speak. Regulars means repeated encounter, which is the substrate of attraction. Low profile means the place is cheap enough to linger in. Playful mood means signals can be sent without commitment. Home away from home means the relationship to the place itself is durable, so the people you meet there can re-encounter you.
These conditions are unusual. Most commercial spaces violate at least half of them. The contemporary American coffee shop is loud, optimizes for laptop revenue, has no regulars because the staff turns over every six weeks, and is explicitly hostile to conversation that would interfere with other patrons' work. The contemporary American bar is optimized for high-velocity revenue, which means loud music to suppress conversation, expensive drinks to discourage lingering, and no seating to maximize standing density. Neither functions as a third place in Oldenburg's sense. The third place has not been replaced by something equivalent. It has been replaced by spaces that look superficially similar but perform an entirely different function.
The historical data on where couples met aligns precisely with this. Rosenfeld's longitudinal series shows bars and restaurants holding remarkably steady as a meeting channel through most of the twentieth century — around 20% — until they begin to collapse in the 2000s. "Meeting at church" was a major channel through 1980, then declines steadily as religious attendance falls. "Met as neighbors" was significant when neighborhoods were dense, walkable, and had front porches; it collapses as suburbanization isolates households behind garage doors. Each line on Rosenfeld's chart corresponds to a category of third place, and each line falls as that category of third place is destroyed by some combination of economic, technological, and zoning pressure.
The collective frame is essential here, because no individual can produce a third place. Third places are emergent properties of a built environment plus a set of social norms plus a price structure. You cannot decide to make Starbucks a third place by going there more often. The economics of the chain will not permit it. What can be done collectively — and is being done, slowly, in some cities — is the deliberate reconstruction of third-place conditions: zoning for mixed-use neighborhoods, supporting independent businesses with rent stabilization, building plazas with free seating, encouraging the small institutions (clubs, leagues, congregations) that depend on third places to function.
What is at stake romantically is the supply of low-stakes, high-frequency encounters between strangers. The apps have substituted high-stakes, low-frequency encounters between strangers: a date is a major commitment of time and emotional load, and a single date with no follow-up costs more attention than ten casual conversations at the same neighborhood bar over six months would have. The third place produced romance as a by-product of community. The app produces romance as a primary task with no community substrate. People do not flourish in environments where every encounter has to mean something. The third place permitted not-meaning, which is the precondition of eventual meaning.
If you take the third place seriously as romantic infrastructure, the implications are unsettling for both the dating-discourse left and the dating-discourse right. The progressive frame says the problem is the apps and we should design better ones. The traditional frame says the problem is moral decline and we should return to courtship norms. Both miss that the precondition for either solution is a built environment that hosts strangers and permits them to linger near each other across many weeks. Neither better apps nor stricter norms can produce that environment. Only zoning, capital, and time can.