There is a particular kind of institutional grief that nobody names at the time. A long-serving nurse retires, or a production manager takes redundancy, or a team is quietly restructured after a merger, and with them goes something that had no line on any balance sheet — a feel for when the machine was about to fail, a memory of why a particular client relationship needed handling with unusual care, a knowledge of which shortcut through the system worked and which ones caused problems three months down the line. I have sat in enough post-merger workshops and strategy away-days to have watched this happen in real time, and what strikes me each time is not the loss itself but the organisation's almost total unawareness that anything has been lost. The filing systems are intact. The org chart has been updated. The new leadership team has a hundred-day plan. And somewhere in all of that confident reconfiguration, something essential has quietly left the building, and no one is quite sure what it was or when exactly it went.

The thing Simone Weil understood about work that most management theory does not

Simone Weil spent eleven months, between 1934 and 1935, working on the factory floor at Alsthom and then at a Renault plant in Boulogne-Billancourt. She was not researching a report. She was trying to understand, in her body, what industrial labour did to a human being's capacity for thought. What she wrote afterwards, collected in the volume translated as The Need for Roots and in her factory journal, is some of the most precise phenomenology of skilled work I have ever encountered. She noticed that the factory system was not simply extracting labour — it was systematically destroying the conditions under which workers could think about what they were doing. The rhythm of the line, the narrowing of tasks, the removal of any feedback loop between the worker and the product — all of it conspired, she argued, to make attention impossible. And without attention, no knowledge is made. It is simply performed, and then forgotten. What Weil understood, and what most management theory has never quite caught up with, is that institutional knowledge does not live in documents or systems. It lives in the quality of attention that people are able to bring to their work. Remove the conditions for that attention — through reorganisation, through surveillance, through the relentless pressure to produce outputs rather than to understand processes — and you do not just demotivate people. You begin, slowly and invisibly, to erase what the institution knows.

Three mechanisms of institutional forgetting

In the engagement work I do with organisations — and I am thinking here of a particular eighteen-month piece of work with a regional public services body that was undergoing its third restructure in seven years — I have come to recognise three distinct mechanisms by which forgetting happens, and they are worth naming precisely because they are so rarely named at all. The first is the departure of tacit carriers. Michael Polanyi's phrase 'we know more than we can tell' is by now well-worn, but its implications are still poorly understood in practice. The person who knows which supplier will go the extra mile in a crisis, who remembers why the northern district team needs a different briefing style, who has internalised over fifteen years the precise texture of what good looks like in this particular context — that person's knowledge is genuinely not transferable by handover note. It is embodied. It leaves when they leave. The second mechanism is the destruction of the conditions for learning. This is Weil's contribution, and it operates more slowly and more insidiously than departure. When people are too pressured to reflect, when meetings exist only to produce actions, when the question 'why do we do it this way?' is experienced as obstruction rather than inquiry — the organisation stops generating new knowledge and begins to consume its existing stock. The third mechanism is what I think of as structural amnesia: the habit of treating each reorganisation as year zero, of assuming that the previous configuration was simply wrong rather than that it contained adaptations to real conditions that may not yet be fully understood. Every restructure that begins by dismissing what came before accelerates this amnesia.

What actually carries knowledge through time

The honest answer, which I have arrived at slowly and which I find most leadership teams resistant to hearing, is that knowledge is carried through time primarily by relationships and by stories — not by documentation, not by knowledge management systems, not by induction programmes, and certainly not by the slide decks that get produced in the final weeks before a long-serving person leaves. I have seen organisations invest heavily in SharePoint architectures and 'knowledge capture' consultancies, and I have seen the results: repositories that nobody visits, documents that describe processes as they were supposed to work rather than as they actually did, explicit knowledge that has been carefully preserved while the tacit knowledge that made it useful has evaporated entirely. What actually works — and here I am drawing on specific instances rather than general principle — is the deliberate cultivation of the conditions under which stories can be told and heard. In one organisation I worked with, a departing director of twenty-two years was given not an exit interview but a series of three long, unstructured conversations with the two people who were going to divide her role between them. No agenda. No output document. Just conversation, with the implicit permission to go wherever the memory went. The people who took part described those conversations, a year later, as the most valuable professional development they had received. The knowledge that transferred was not procedural. It was relational and contextual — an understanding of why certain things worked, of what the organisation had tried before, of which external relationships carried weight that was not visible in the formal record.

The role of attention, and why it is the first casualty of growth

Weil's concept of attention — l'attention — is not about concentration in the narrow sense. It is about a quality of receptive, patient, non-grasping awareness that she believed was both the foundation of genuine knowledge and the precondition for ethical relation to another person. She developed this idea most fully in her essay 'Reflections on the Right Use of School Studies with a View to the Love of God,' which is not about education in any institutional sense but about the cultivation of a particular inner posture. I think about this essay often in the context of organisational life, because what it describes — the effort to remain open and present rather than to impose a predetermined framework on what you are seeing — is precisely the capacity that growth and pressure most reliably destroy. When an organisation is small, attention is often abundant. People know each other's work, know the history of decisions, know the texture of the relationships that make things possible. As the organisation grows, as the pressure to standardise and to scale increases, the conditions for that quality of attention narrow. Meetings become denser and shorter. The expectation that people will carry institutional memory gives way to the expectation that systems will carry it. And the systems, of course, can only carry what has already been made explicit — which is to say, they can only carry the part of knowledge that was, in a real sense, already dead.

What leaders can do, and why most of what they try does not work

I want to be careful here not to produce a list, because the problem is not one that admits of listed solutions. But there are orientations that seem to help and orientations that reliably do not. What does not work: knowledge management systems treated as ends in themselves, exit interview processes that ask the wrong questions and produce documents that nobody reads, 'lessons learned' sessions held too quickly after a project closes and too far from the people who actually did the work, and the general assumption that if something is important it will be visible and if it is not visible it is not important. What does work, or at least what I have seen work in specific contexts: creating what I would call 'slow time' in the organisation — protected space where people can think together about what they know and how they know it, where the purpose is not to produce an output but to cultivate understanding. One organisation I worked with introduced what they called 'working history' sessions: quarterly conversations, no more than six people, no agenda, focused only on the question 'what have we learned in the last three months that we did not know before?' Over two years, these sessions became the organisation's primary mechanism for institutional learning — not because they produced documents, but because they built the habits and the relationships through which knowledge could move laterally across the structure.

The ethics of remembering

There is, finally, an ethical dimension to this that I do not want to pass over too quickly. When an organisation forgets what it once knew how to do, the cost is not only operational. The people whose knowledge has been lost — the ones who carried it and were not asked to pass it on, or who were asked in ways that could never have worked — often experience that loss as a form of erasure. I have spoken with people who worked for thirty years in an organisation, who built something real, who understood something about their field that had taken them decades to learn, and who left feeling that none of that was valued because none of it was legible to the systems and metrics through which the organisation measured value. Weil would have recognised this: her factory journal is full of this grief, the grief of workers who were required to perform knowledge without being permitted to possess it. To take institutional forgetting seriously is not only a strategic choice. It is a way of deciding what kind of organisation you want to be — one that treats the knowledge of its people as a resource to be extracted and then discarded, or one that treats it as something worth inheriting.

I do not have a clean resolution to offer here, because the problem is not one that resolves cleanly. What I keep returning to is the word 'inheritance' — the idea that what an organisation knows is not a possession but a stewardship, something received from those who came before and owed, in some form, to those who come after. The organisations I have seen hold their knowledge best are not the ones with the most sophisticated systems. They are the ones where somebody, somewhere, understood that remembering is an act of care.