You're late to a seminar. You cut across the quad that separates the engineering wing from the life sciences block, and you notice, not for the first time, that you are completely alone. Nobody crosses this particular stretch of campus unless they have to. The benches are there. The path is paved. And still, the two buildings face each other across the gap like neighbors who stopped speaking years ago and have simply made their peace with it. The map is not neutral. It is an argument.
The spatial organisation of a research campus is one of the most honest documents an institution produces. More honest, certainly, than its strategy papers. Buildings take decades to build and decades to demolish, which means the intellectual assumptions baked into their arrangement outlast the administrators who commissioned them. Walk the original Rockefeller University campus in New York and you are reading the priorities of early twentieth-century biomedical science: tight clustering, shared corridors, a layout premised on the idea that virologists and biochemists should collide by accident on their way to lunch. The collision was the point. The architecture encoded a theory of how discovery happens.
The corridor as intellectual policy
Consider two researchers: call them Dr. Ama Owusu, a materials scientist, and Dr. Jonas Breit, a synthetic biologist. Same institution. If their departments share a building with a common coffee point on every floor, Ama and Jonas will, statistically, meet. They'll complain about grant deadlines, notice overlapping problems, and perhaps eventually write a joint proposal. If their buildings are separated by a ten-minute walk across an exposed car park, that meeting never happens. Not because either of them is incurious, but because friction is real and time is short. The campus layout has quietly decided, on their behalf, that materials science and synthetic biology are not related fields.
This is not a trivial administrative choice. Research on scientific collaboration consistently finds that physical proximity correlates with co-authorship at rates that drop off sharply beyond about fifty metres between offices. Fifty metres. That is the distance across which a campus planner can accidentally determine whether an interdisciplinary breakthrough happens or whether two people simply never learn the other exists.
Older campuses, built when disciplines were younger and boundaries more porous, often show a different logic. The original University of Edinburgh medical quarter placed anatomy, chemistry, and natural philosophy within easy walking distance, not because anyone theorised about interdisciplinarity, but because the Enlightenment university hadn't yet decided these were separate enterprises. The separation came later, institutionalised in the Victorian period as specialisation became both intellectually fashionable and bureaucratically convenient. The buildings followed. Departments got their own front doors, their own entrances, their own gravitational fields. The campus began to look like a small town whose neighbourhoods have forgotten how to mix.
The twentieth century produced two competing responses to this fragmentation. One was the mega-building: a single structure large enough to house multiple departments, with shared atria, open staircases, and seminar rooms bookable by anyone. The Salk Institute in La Jolla, designed by Louis Kahn, is the canonical example. Its two flanking laboratory wings face each other across an open courtyard with a channel of water running to the Pacific horizon. Nothing about it is accidental. Kahn and Jonas Salk argued at length about whether the design would produce the right kind of thinking. Salk wanted scientists to feel they were working at the edge of the known world, and the architecture was meant to provoke exactly that vertigo, the sensation of standing at a threshold with nothing settled behind you.
The other response was the research park model: disperse everything, give each group its own building, connect them with pleasant paths and a shuttle bus. This approach tends to produce exactly what you'd expect from pleasant paths and shuttle buses. People stay put.
What people get wrong about the open-plan fix
The fashionable correction, especially in tech-adjacent campuses built in the years after open-plan offices became a Silicon Valley article of faith, has been the hot-desking breakout zone, the deliberately engineered serendipity machine. The instinct is right. The execution, almost everywhere I've seen it tried, is not. Forced proximity without intellectual permission to cross disciplinary lines produces noise, not collaboration. A physicist who has been told, implicitly or explicitly, that her promotion case will be judged entirely on physics publications has no career incentive to spend three hours talking to a sociologist in a shared atrium, however beautifully the atrium is lit. The building can invite the conversation. It cannot fund it, tenure it, or protect it from a departmental review committee.
And here is the question worth putting to any institution that has recently spent money on a gleaming new interdisciplinary hub: did you also change how you hire, how you promote, and whose grants count toward whose targets? Because if you didn't, the building is a stage set. Open staircases, empty between performances.
This is the gap that campus designers and university administrators almost never close together. The spatial argument and the incentive structure have to agree, and the honest truth is that most institutions find it far easier to commission an architect than to renegotiate a tenure committee's criteria. Architecture is expensive but legible. Institutional culture is cheap to describe and very nearly impossible to shift.
The most revealing thing you can do on any research campus is find the oldest, shabbiest corridor nobody has renovated yet. That is where the institution's original theory of knowledge is still legible, written in brick and floor plan, like a palimpsest the paint budget never quite reached. It tells you what the founders thought counted as a discipline, who was meant to talk to whom, and which fields were considered too different to share a wall. Some of those assumptions were right. Most of them just aged in place, quietly shaping what got discovered and what didn't, and the tragedy is that nobody had to make that choice consciously. The buildings made it for them.