Institutions inherit far more than they deliberately choose. They inherit laws written for earlier problems, technologies selected under previous constraints, organisational structures created by past reforms, professional routines developed by earlier generations and categories that reflect ways of understanding the world that may no longer seem obvious today. They inherit buildings, contracts, databases, jurisdictions, relationships, expectations and memories. By the time the people currently responsible for an institution begin making decisions, much of the architecture within which those decisions will be made already exists. The present institution is always, to some degree, an inheritance.

This can make institutional change appear deceptively simple from the outside. If a procedure no longer works, change it. If an organisational structure has become inefficient, redesign it. If a technology is obsolete, replace it. Yet institutions rarely encounter their inherited arrangements as isolated choices waiting to be reconsidered. Existing structures are connected to other structures. Procedures have accumulated expectations around them. Technologies have generated complementary systems. Professional groups have developed expertise suited to particular arrangements. Budgets, legal mandates and authority relations have adapted to what already exists. What began as one choice can gradually become the environment within which later choices must be made.

This is the basic intuition behind path dependence. Where an institution can go is partly shaped by where it has already been. Earlier decisions alter the costs, possibilities and meanings of later ones. Once an institutional path has developed, changing direction may require more than simply deciding that another arrangement would now be preferable.

But there is a danger in taking this insight too far. If history constrains every institutional choice, it can begin to sound as though history determines every institutional future. Existing arrangements appear inevitable because alternatives have become difficult to imagine. Practices survive because they are embedded, and their embeddedness becomes evidence that they must survive. The institution gradually stops distinguishing between what it inherited and what it genuinely requires.

That is the point at which path dependence risks becoming historical determinism.

The distinction matters because what exists because history produced it is not the same as what must continue to exist because history produced it. An institution may have good reasons for preserving an inherited arrangement. A rule may still protect an important value. A procedure may embody hard-won experience. A legacy system may remain the most practical option given current constraints. Historical agency does not require treating inheritance as a mistake. It requires something more modest but more consequential: recognising that inherited structures have histories, and that having a history makes them explicable rather than inevitable.

Imagine an organisation whose approval process requires several layers of review. Employees may describe the process simply as “how approvals work”. Yet perhaps those layers accumulated at different moments: one after a financial scandal, another during a period of political scrutiny, another when authority was reorganised and a final one when a new information system was introduced. Each layer may once have solved a real problem. Over time, however, the reasons disappear from everyday awareness while the structure remains.

When the history becomes invisible, the architecture can begin to look natural. People no longer see four historical responses to four particular circumstances. They see “the approval process”. The distinction is subtle but important. An institution loses part of its agency when contingent historical outcomes become cognitively indistinguishable from permanent necessities.

This does not mean that simply discovering the history makes change easy. Some inherited arrangements are extraordinarily difficult to alter. They may be embedded in law, infrastructure, professional identity or public expectations. Investments may be irreversible. Organisational systems may depend on one another so extensively that changing one component creates costs elsewhere. Political legitimacy may itself be attached to institutional continuity. Understanding that an arrangement is contingent does not magically remove these constraints.

Historical agency therefore should not be confused with freedom from history. An institution cannot return to an imaginary blank slate and choose again without consequences. Sunk costs remain sunk. Previous commitments have shaped capabilities. People have built careers and expectations around existing structures. Earlier decisions continue to exert causal force even after the institution understands where they came from. Recognising contingency changes what the institution can think about its architecture; it does not erase the architecture.

This is why institutional evolution begins cognitively before it becomes organisational. Before an institution can treat its own form as something that might evolve, it must be able to perceive a difference between inheritance and necessity. It must be capable of asking why a structure exists, what problem originally produced it, which conditions have changed and which constraints remain genuinely binding. Without that distinction, the future is interpreted through the architecture of the present, and the architecture of the present is interpreted as though history had no alternatives.

The problem appears in reform repeatedly. Institutions sometimes attempt to change visible structures while leaving inherited assumptions untouched. A new technology is inserted into an old workflow. A new organisational unit inherits the categories of the unit it replaced. A reform changes formal responsibilities while preserving the same information flows and authority relationships. The institution appears to move while remaining cognitively attached to the path that produced its previous form.

Historical agency begins when that attachment becomes observable.

This is not yet a theory of institutional redesign. Recognising that an arrangement is historically contingent does not tell us which alternative should replace it, who has legitimate authority to make that choice or how competing interests should be reconciled. Those are different questions. Historical agency begins one step earlier: with the capacity to see that the inherited institution is one historically produced configuration rather than the only configuration that could exist.

That capacity matters because institutions cannot deliberately evolve while treating their current form as inevitable. Learning requires the possibility that previous arrangements can be reconsidered. Adaptation requires distinguishing constraints that remain real from constraints that merely survive. Institutional memory becomes more valuable when it preserves not only what happened but why structures were created. Even apparently permanent features of an organisation become cognitively different once their provenance can be examined.

There is also a deeper reason this matters for institutional intelligence. An institution that cannot distinguish historical inheritance from necessity may continue solving problems that no longer exist while remaining poorly equipped for problems that do. Its past becomes not a source of learning but a silent author of the future. The institution continues changing, because all institutions change, but the direction of change remains heavily governed by structures whose historical origins have disappeared from view.

The alternative is not an institution without history. Such an institution is impossible and would probably be undesirable. Historical inheritance carries accumulated knowledge, legitimacy, experience and stability as well as constraint. The goal is therefore not to escape history but to become capable of relating to it differently: to recognise what has been inherited, understand why it emerged and distinguish the constraints that still matter from arrangements whose inevitability exists mainly because their contingency has been forgotten.

Institutional evolution begins when history remains influential without remaining destiny. An institution acquires historical agency not when it becomes free from everything that shaped it, but when it can see its inherited architecture clearly enough to recognise that the past has produced the present without acquiring an automatic right to determine the future.