We talk about institutions as if they were things we could point to. A ministry occupies a building. A university has campuses. A hospital has wards, employees and equipment. A public agency has an organisational chart, a legal mandate, databases, procedures and a workforce. Yet none of these answers a surprisingly difficult question: where, exactly, is the institution? If the building closes and the organisation moves elsewhere, we normally say that the institution continues. If its databases are replaced, we do not say that a new institution has suddenly appeared. Even its organisational chart may be redesigned without destroying the institution itself. The visible components matter, but none of them seems sufficient to tell us what the institution actually is.
The problem becomes clearer if we imagine a more radical change. Suppose that, over time, every person working in an organisation leaves and is replaced. The ministers change, the civil servants retire, new specialists arrive, managers move elsewhere and an entirely different generation eventually occupies every role. Would we necessarily say that the original institution had disappeared? Usually, we would not. We would still speak of the same ministry, the same university or the same agency, even though none of the people who once constituted it remained. This simple thought experiment reveals something important: institutional continuity cannot be identical to the continuity of institutional members.
Something else must persist. Authority can be reproduced as one office-holder replaces another. Routines can survive the people who first performed them. Records allow previous decisions to influence later ones. Roles specify responsibilities that continue even when their occupants change. Procedures, expectations and established relationships teach newcomers how the organisation operates. Institutional memory may be imperfect, and institutional arrangements may gradually change, but enough of this architecture can persist for the institution to remain recognisable across successive generations of people.
This does not mean that an institution is an invisible object floating somewhere above its members. Institutions still depend on people, material resources, technologies and formal structures. The point is different: the institution is not exhausted by any one of those components. What persists is partly the organised pattern through which components are connected and repeatedly brought into action. A person may leave, but a role remains. A document may be replaced, but the procedure that requires a certain kind of record may continue. A leader may change, but an authority relation may survive. The institution exists through this capacity to reproduce an organised pattern while its particular components change.
That distinction becomes especially important when we begin to speak about cognition. A group of people can think together without constituting an institution. A temporary crisis team may combine expertise, exchange information and reach a judgement that no member could have reached alone. A scientific workshop may generate a collective insight. A group of strangers may coordinate extraordinarily well for a few hours. These are genuine forms of collective cognition, but they are not automatically institutional cognition. Collective cognition and institutional cognition are not the same thing.
What an institution adds is persistence. Its cognitive arrangements can outlast the particular people involved in a single episode. It also adds authority: some observations, judgements and decisions have institutional standing while others do not. It adds memory: previous interpretations and choices can be preserved and made available to later actors. And it adds reproducibility: similar patterns of attention, coordination and decision can recur because the architecture that supports them continues to exist. A temporary group may think together once. An institution can reproduce conditions under which cognition happens again and again.
This is why asking where an institution is turns out to be more than a philosophical curiosity. If we imagine the institution simply as the people currently inside it, then institutional performance appears to depend mainly on the intelligence, expertise or judgement of those individuals. When performance is poor, the natural response is to replace people, train them better or recruit more capable ones. Sometimes that is exactly what is required. But if institutional properties also reside in persistent arrangements of authority, memory, routines and relationships, then changing personnel may leave the deeper architecture untouched. New people can enter an old pattern and begin reproducing many of the same outcomes.
The reverse is also possible. An institution may preserve its name, legal mandate and much of its workforce while changing profoundly because the architecture through which it operates has changed. Information may reach different people. Authority may be redistributed. Memory may become more accessible. Previously isolated roles may begin to coordinate. New routines may alter what is noticed, questioned or remembered. From the outside, many of the visible components may remain familiar, while the institution’s way of operating has become substantially different.
Thinking in these terms gives us a more useful starting point for understanding institutional performance. Before attributing success or failure entirely to the people currently occupying an organisation, we need to look at the architecture through which those people operate. An institution is not simply the sum of its present members or resources, but a persistent and reproducible arrangement of authority, routines, memory and relationships that can survive changes in the particular components through which it is expressed. Once that architecture becomes visible, many problems that appear to be failures of individual competence can be recognised as something different: failures in the institutional conditions that shape what competent people are able to see, remember, coordinate and do.
This changes the practical question as well. Improving an institution cannot always mean simply adding better people, more information or new technology. Sometimes improvement requires changing the arrangements that allow knowledge, authority, memory and human capability to persist and work together over time. Once we stop looking for the institution in any single person, building, database or organisational chart, we can begin to see institutional performance differently: not merely as the sum of individual capacities, but as an outcome of the architecture that enables those capacities to become durable, coordinated and institutionally consequential.