A prison exists to punish. A monastery exists to devote. A psychiatric hospital exists to treat. A military barracks exists to train. A boarding school exists to educate. Ask what these five buildings are for and you get five unrelated answers. In 1961 the sociologist Erving Goffman asked a different question — not what they're for, but how they're built to run day to day — and found the same answer underneath all five. He called it the total institution, and the "total" is doing precise work in that name: it's a place where every part of a person's life happens inside one set of walls, under one authority, with no separation left between where you sleep, work, eat and are watched.
Four features, not one
Goffman's definition rests on a short list of shared traits, and it's the combination that matters, not any single one alone. All aspects of life happen in the same place under the same single authority — sleep, work, and leisure aren't spread across different institutions the way they are outside, they're all inside this one. A large batch of similarly situated people are handled together as a group rather than as individuals, moved through the same schedule at the same time. That schedule is imposed from above in tight, explicit detail, with each activity's timing decided for the group by the staff rather than negotiated by the people living it. And a sharp, visible split separates the "inmates" — Goffman's own deliberately blunt word, meant to apply equally to a patient, a recruit, a novice or a prisoner — from the staff who run the place and go home at the end of the day.
New arrivals typically go through what Goffman called a mortification of self: personal clothing exchanged for a uniform, a name replaced or supplemented with a number or rank, possessions confiscated or restricted, searches and inspections that leave no part of the incoming person still private. The purpose isn't cruelty for its own sake — a functioning total institution needs everyone legible and interchangeable to the staff running it, and the previous civilian self is the first thing standing in the way of that.
The same shape, doing opposite jobs
What makes the theory worth having isn't that these five places are secretly the same institution — they plainly aren't, and living in one is a wildly different experience depending on which. It's that the shared structure predicts real, recurring behaviour independent of the institution's stated purpose: the same "underlife" of small, unauthorised adaptations — hoarded contraband, private jokes at the staff's expense, informal favour economies — appears in prisons and psychiatric wards and monasteries alike, because the structural pressure producing it (total control, total surveillance, an us-versus-them split) is the same pressure in each case, whatever the institution officially exists to do about it.
What we're still unsure about
Goffman's fieldwork leaned heavily on one mid-20th-century American psychiatric hospital, and applying a theory built there to prisons, monasteries and boarding schools involves real generalisation his original data didn't fully test. The framework has also drawn a specific, fair criticism: it treats consent almost as an afterthought, describing a monk who joined voluntarily and a prisoner sentenced against their will with much of the same vocabulary, when whether someone can walk out the front door is arguably the most important variable of all, not a footnote to a shared "shape."
This sits inside Total Institutions (Goffman), one of seven topics in Institutions, one of four domains in Sociology, one of seventeen subjects the app can quiz you on.