There is a piazza in Lucca, Italy shaped like an ellipse. The ellipse has been there for nearly 2,000 years. The building that made it has not.
A Roman amphitheater occupied that site in the 2nd century AD. Over the medieval centuries, inhabitants quarried its stone for other construction. As each ring of seats came down, buildings went up where the stone had been. Property lines followed the old seating rows. Streets followed the old corridors. The amphitheater dissolved into the city's fabric and left its geometry behind. An architect named Lorenzo Nottolini formalized the space in 1830, clearing the interior so people could use what had been there all along.
The building is gone. The form persists.
This is what cities do that buildings cannot. They remember.
Types vs. Programs
In 1306, an Augustinian friar named Fra Giovanni added a single unified roof to a civic hall in Padua that had been under construction since 1172. The building had already hosted courts, markets, and civic ceremonies for over a century. It kept hosting them. Then concerts. Then exhibitions. Today it houses a weekly market. Over 800 years, no single function has defined it. Its dimensions — 81.5 meters long, 27 meters wide, open span — are what made it flexible enough to absorb any use that fit the pattern.
This is what Aldo Rossi called a type: a structural pattern that encodes a solution to a recurring human need, in a form general enough to outlast any specific program.
The courthouse-market-meeting hall is a type. The courtyard house is a type. The basilica is a type — it was a Roman law court before it was a church, and both uses expressed the same form: nave, aisles, apse, hierarchical entry.
A program is what a building needs to do right now. A type is what the building can hold over centuries.
The 20th century built programs and called them architecture.
What Urban Renewal Actually Erased
The Housing Act of 1949 launched the American urban renewal program. By 1974, it had demolished approximately 600,000 housing units. The program targeted what planners called "blight" — a designation based on building age and residential density. It rarely measured what those buildings contained.
Herbert Gans spent years inside the West End of Boston before the city demolished it in 1958. He published his findings in The Urban Villagers while the demolition was still happening. By every social measure — crime rates, employment, community organization, social cohesion — the neighborhood was functioning. The city classified it as blighted based on its age.
In St. Louis, Pruitt-Igoe replaced the DeSoto-Carr neighborhood with 33 high-rise towers in 1954. The first towers came down in 1972. Total operational lifespan: 18 years. The neighborhood it replaced had stood for a century.
The planners' models measured blight by looking at buildings. They didn't have a column for what those buildings remembered.
The corner bar where three generations drank after work. The market square where the same vendors had sold to the same families for decades. The street where everyone knew which windows to watch. This is not sentiment. It is the physical infrastructure through which communities accumulate the social trust that makes ordinary life possible.
You can build a new bar in the replacement development. It will serve drinks. It will not have memory. No one in it has a history of being there that predates their own biography.
The Same Mistake, Different Domain
In the 1840s, most of Ireland's potato crop grew from a single variety: the Lumper. Before agricultural consolidation, Irish farmers cultivated dozens of distinct varieties, each adapted to local soil and climate. Consolidation simplified that diversity in favor of the highest-yielding option.
One pathogen found the Lumper. The crop failed entirely.
The urban renewal planners made the same error. Traditional cities maintained diverse typologies — the market hall, the courtyard house, the arcade, the workshop neighborhood, the residential square. Each was adapted to different uses, different social patterns, different relationships between public and private life.
International Style modernism replaced them with variations on one type: the freestanding tower with identical floor plates, separated from other uses by zoning law.
When the program failed — when crime rose, when social isolation set in, when the maintenance model broke down — every tower failed in the same way. There was no resilient variety to fall back on.
The type library is the ecological repertoire. Wipe it out and you get the same fragility.
What Putnam Measured, What Rossi Explained
Robert Putnam documented the collapse of American social capital across the second half of the 20th century. Declining civic participation. Declining trust in neighbors. Declining informal social contact. He correlated it with suburbanization, television, and the replacement of walkable neighborhoods with car-dependent sprawl.
The mechanism is physical. Social capital is built through repeated encounters in shared spaces with shared histories. Those encounters require places that carry meaning prior to your own arrival — where the weight of previous use creates a reason to be there.
Urban renewal didn't just clear buildings. It cleared the accumulated reasons to gather. The new plaza has no prior claim on you. The new development's bar is not the bar your grandfather's friend ran. You have no history with it. It has no gravitational pull.
Putnam measured the social capital deficit. Rossi provides the mechanism: when you destroy collective memory, you destroy the material basis for collective identity. Identity needs substrate. Substrate requires time.
The Patience Problem
The amphitheater in Lucca didn't survive because anyone preserved it. No one decided to keep the ellipse. The medieval inhabitants were not historians. They quarried the stone and built their houses where the seats had been. The ellipse survived because form was more durable than any individual decision about it.
That's the difficulty. You cannot engineer collective memory. You can only refrain from destroying what has already been accumulated.
The Colosseum's games ended in 523 AD. It has been a quarry, a fortress, a shrine, and a monument in the 1,500 years since. Functionalists would have cleared it in 524. The city held it anyway, in the slow way cities hold things: by keeping the form long enough that the next use could find it.
The question isn't whether we can build places with memory. We can't. Memory isn't installed. It accretes. The question is whether we have the patience to stop building programs in places where accumulated form has been waiting for the next thousand years.