What an Institution Forgets When Its People Leave

Safety Lives in Habits, Not in Files

Why do institutions exist at all? One answer is that they let us forget safely. A well-run organisation remembers so that no single person has to: where the bottleneck forms, what rain does to a crowd, which announcement causes panic. When that memory is lost, an institution can look the same from outside and yet be dangerously different inside.

A puja that had done this ninety times Chorbagan's Durga Puja was in its 91st year.

Nearly a century of experience should have prepared it for a rainy evening. On 17 October, a sudden downpour sent visitors rushing towards the pandal for shelter. Witnesses describe a narrow entrance, an equally narrow exit and an unlit lane.

Early official information suggests a hurried announcement told people to leave because the pandal might collapse. Nineteen people died. None of these risks is new. Prateek Raj, who has studied crowd management at the Kumbh Mela, argues that such mega events succeed because organising knowledge is spread across many grassroots hands: spillover space, multiple entries and exits, calm and timely communication, volunteers stationed in the right places.

Experienced organisers do not consult a rulebook for this. They remember. That is why one detail matters. The committee changed leadership this year; its former general secretary, who had served 15 years, resigned in July.

Whether the old team's operational know-how passed to the new one is a question only an independent inquiry can answer. But the possibility alone shows how fragile unwritten knowledge is. Mumbai saw the same pattern in 2017.

At Elphinstone Road station, sudden rain packed commuters onto a narrow footbridge, a rumour of collapse spread and 23 people died. Rain, a bottleneck and a panicked message: the same three ingredients.

Rockets, too, can be forgotten The theme is not limited to crowds.

NASA's F-1 engines powered the Saturn V rockets that carried astronauts to the Moon. As is often recounted, when engineers decades later sought to study and recreate the engine, the blueprints had survived, but much of the practical expertise had not.

Unrecorded workarounds and hands-on judgment had retired with the engineers who built it. The contrast is instructive. NASA had documents but lost practice. A neighbourhood puja has practice but few documents. Each fails in its own direction: one cannot use what it wrote down; the other cannot pass on what it never wrote.

What the

comparison shows Put together, the two cases reveal something neither shows alone. Safety knowledge lives in practice, in people doing the work together and paper alone cannot hold it. So continuity cannot be left to chance or to goodwill.

It has to be designed: overlapping terms for outgoing and incoming leaders, rehearsals before every season and the habit of writing down what only experience teaches. For the Essay paper, the lesson is that institutions are not buildings or titles.

They are memory in motion and memory survives only when an organisation deliberately teaches it to its next generation.